UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A study of variation in the genus Alaria Greville Widdowson, Thomas Benjamin 1964

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-UBC_1964_A1 W5.pdf [ 11.57MB ]
JSON: 831-1.0105799.json
JSON-LD: 831-1.0105799-ld.json
RDF/XML (Pretty): 831-1.0105799-rdf.xml
RDF/JSON: 831-1.0105799-rdf.json
Turtle: 831-1.0105799-turtle.txt
N-Triples: 831-1.0105799-rdf-ntriples.txt
Original Record: 831-1.0105799-source.json
Full Text

Full Text

A STUDY OF VARIATION IN THE GENUS ALARIA G-REVILLE  by  Thomas Benjamin Widdowson .B.A., The University of British Columbia, 1957 M.A., The University of British Columbia, 1959  A. thesis submitted in partial fulfilment, of the requirements for the degree of Doctor of Philosophy in the Department of Biology and Botany We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA April, 1964  In presenting this thesis i n p a r t i a l fulfilment of the requirements for an advanced degree at the University of B r i t i s h Columbia, I, agree that the Library shall make i t available for reference and study-  freely  I further agree that per-  mission for extensive copying of this thesis for scholarly purposes may be granted by the Head of my Department or by his representatives.  It i s understood that, copying or publi-  cation of this thesis for financial gain shall not be allowed without my written permissions  Department of The University of B r i t i s h Columbia, Vancouver 8, Canada Date  The University of British Columbia FACULTY OF GRADUATE STUDIES PROGRAMME OF THE FINAL ORAL EXAMINATION FOR THE DEGREE OF DOCTOR OF PHILOSOPHY of THOMAS BENJAMIN WIDDOWSON B.A., The University of British Columbia, 1957 M.A., The University of British Columbia, 1959 IN ROOM 3332, BIOLOGICAL SCIENCES BUILDING MONDAY, JANUARY 27, 1964,  at 4:00 P.M.  COMMITTEE IN CHARGE Chairman: F.H. Soward B.M. Bary Kathleen Cole V . J . Krajina S.W. Nash  G.L. Pickard R.F. Scagel Janet R. Stein T.M.C. Taylor  External Examiner; A.R.A. Taylor University  of New Brunswick  A STUDY OF VARIATION IN THE GENUS ALARIA ABSTRACT Field collections and observations of Alaria (Order Laminariales) were made throughout the intertidal zones of the temperate areas in the Northern Hemisphere. In addition a study was made of a l l available herbarium material. Morphology was chosen as the basis of the systematic study of the genus. The systematics of the genus were studied by. a discriminant and distance func= tion analysis, using an IBM 1620 computer. From this analysis, the characteristics of 10 species were defined. Four other species were described from herbarium material, but were not sampled in sufficient quantity for statistical analysis. The dividing lines drawn between the various species are meaningful but essentially arbitrary. Confusion in the faxonomy of the genus has two main sources. F i r s t , the taxa are not completely differentiated into distinct species. Second, differences of environment appear to play a greater role as a cause of morphological variation than do differences of genotype. The names of 107 specific and subspecific taxa of the genus Alaria, including recombinations, were dis= covered in an exhaustive search of the literature. Of these 107 names, 19 were eliminated as illegitimate under the rules of the International Code of Botanical Nomenclature, or were removed from the genus. The genus Pleuropterum Miyabe et_ Nagai-was reduced to synonomy under the genus Alaria. The conservation of the generic name Alaria was re-examined and found to be adequate. A thorough search was made for possible holotypes or lectotypes of the 88 taxa remaining. Material was either shown to be the holotype or found and designated as the lectotype for 66 taxa. No relevant material could be found for 11 more of these taxa and possible material for another 11 was judged inadequate for pract i c a l systematic purposes.  GRADUATE STUDIES Field of Study:  Botany  Biological Oceanography Marine Phyeology Marine Phytoplankton  R.F. Scagel R.F. Scagel R.F. Scagel Janet R. Stein  Related Studies; Biology of Fishes Biophysics Chemical Oceanography Dynamic Oceanography Marine Zooplankton Synoptic Oceanography TaJtonomy of Vascular Plants  C.C. Lindsey 0. Bliih J.D.H. Strickland G.L. Pickard R.F. Scagel G.L. Pickard T.M.C. Taylor  ii ABSTRACT Field collections and observations of the brown alga Alaria Greville (Order Laminariales) were made throughout the intertidal zones of temperate areas in the Northern Hemisphere. In addition, a study was made of a l l available herbarium material Morphology was chosen as the basis of the systematic study of the genus.  The systematics of the genus were studied by a discriminant and dis-  tance function analysis, using an IBM 1620 computer. characteristics of ten species were defined.  From this analysis the  These ten species are;  Alaria  angusta. 4. crassifolia. A. crispa. A. esculenta. 4 . fistulosa. A. marginata. A. nana. 4 . r>raelonga. A. taeniata. and 4 . tenuifolia.  Four other species  were described from herbarium material, but were not sampled by the writer i n sufficient quantity for s t a t i s t i c a l analysis. These four species are: 4 . grandifolia. 4 . octac-tensis, A., paradisea, and A. p y j a i i . The dividing lines drawn between the various species are meaningful but essentially arbitrary. sources.  Confusion in the taxonomy of the genus has two main  First, the taxa show intergradations that make their delimitation  difficult.  Second, .differences of environment appear to play a greater role  as causes of morphological variation than do differences of genotype. The names of 107 specific and subspecific taxa of the genus Alaria. including recombinations, were discovered in an exhaustive search of the literature.  Of these 107 names, 19 were eliminated as illegitimate under  the rules of the International Code of Botanical Nomenclature., or were removed from the genus.. The genus Pleuropterum Miyabe et. Nagai was reduced to synonymy under the genus Alaria.  The conservation of the generic name -of  iii A l a r i a was re-examined and found t o be adequate. A thorough search was made f o r possible holotypes or lectotypes of the 88 taxa remaining. M a t e r i a l was e i t h e r shown t o be the holotype or found and designated as the lectotype f o r 66 taxa.  No relevant m a t e r i a l could  be found f o r 11 more of these taxa and p o s s i b l e m a t e r i a l f o r another 11 was judged inadequate f o r p r a c t i c a l systematic purposes.  - iv TABLE OF CONTENTS  Introduction — scope of the problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . -- h i s t o r i c a l background — approach followed i n t h i s study -— — --  1 2 3  d e f i n i t i o n of measurements . . . . . . . . . 7 . . . . . . . . . . . . . . . . . . . 5 pattern of growth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 d i s t r i b u t i o n and ecology of the genus . . . . . . . . . . . . . . . . . . 9 h i s t o r i c a l changes i n d i s t r i b u t i o n 12  —- l i s t of names discussed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Statistics  Systematics  — the discriminant and distance functions  23  — f i r s t run  29  — fourth run — use of the discriminant and distance function . . . . . . . . . . -- decisions based on the s t a t i s t i c a l analysis . . . . . . . . . . . .  36 37 38  — — — — — — — — — — — — — — — —  limits key Alaria Alaria Alaria Alaria. Maria Alaria Alaria Alaria Alaria Alaria Alaria Alaria Alaria Alaria  of the genus angusta crassifolia crispa . . . . . esculenta fistulosst grandifoiia marginata nana pchotensis paradisea nraelonga pylaii taeniata tenuifolia  ............  41 43 45 47 49 51 55 57 59 62 -64 65 66 68 70 71  Discussions and Conclusions eva3.uation of techniques — possible causes of v a r i a t i o n — e f f e c t s of environment on d i s t r i b u t i o n — possible coursecdf evolution  73 75 79 80  Bibliography  85  •  _ v APPENDIX I 1. 2.  — Nomenclature.  Discussion of typification and legitimacy of names Synopsis of taxonomic criteria  93 115  APPENDIX II — Statistical data. Table Table Table Table Table  1. — Deseritptive data of samples used 2 . — Data;concerning duplicate samples 3. —Distance function matrices for f i r s t and second runs 4. — Results of third test or Run 3 5 . — Technical data from the fourth run APPENDIX III — Geographical  118 121 122 124 125  distributions.  Table 1. —Additional localities for the species of Alaria 127 Table 2 . — Position of localities cited elsewhere than Table 1 above . . 135 LIST OF FIGURES Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure  1. — 2. - 3. — k. — 5• " 6. —  Illustration of morphological terminology Patterns of growth in Alaria esculenta Rates of growth in three Alaria species Type localities and geographical regions Some presence and absence factors Oceanographic features of the southern limit of Alaria in California . . 7. — Oceanographic features of the southern limit of Alaria in New England 8. — Oceanographic features of the southern limit of Alaria in the English Channel 9. Oceanographic features of the southern limit of Alaria in Japan 10.— Lower limits of Alaria in the subtidal zone 11.— Beds of Alaria fistul.osa in southeast Alaska 12.— Distribution of Alaria in southwest England 13.— Results of the f i r s t run . Ik.— Frequency distribution of N/D values 15.— Results of the second run 16.— Relationships between the Atlantic and the Pacific samples 17.— Results of the fourth run 18.— Distribution of Alaria anpusta 19.— Distribution of Alaria crassifolia 20.— Distribution of Alaria crista 2 1 . — Distribution of Alaria esculenta 22.— Distribution of Alaria flstulosa 2 3 . — Distribution of Alaria grandifolia 2k.— Distribution of Alaria marginata 25.— Distribution of Alaria nana 26.— Distribution of Alaria ochotensis 2  140 lk2 ikk 1^6 1^8 150 152 15^ 156 158 l6o 162 l6k 166 168 170 172 175 177 179 l 8 l 183 185 187 189 191  - vi Figure 27 Figure 28.-Figure 29.-Figure 30.-Figure 31. -Figure 32.-Figure 3 3 . -Figure 3 4 . -Figure 3 5 . - Figure 36.-Figure 37.-Figure 38.-Figure 39.-Figure 4 0 . -Figure 4 1 . -Figure 4 2 . -Figure 4 3 . -Figure 4 4 . -Figure 4 5 . -Figure 4 6 . -Figure 47.-Flgure 4 8 . -Figure 4 9 . -Figure 50.-Figure 51.-Figure 52.-Figure 5 3 . -Figure - 5 4 . -Figure 5 5 . Figure 56.Figure 5 7 - ~ Figure 58.Figure 59 Figure 60.Figure 6 l . -  Distribution of Alaria paradisea . . . . . . . . . . . . . . . . . . . . . Distribution of Alaria praelonga . . . . . . . . . . . . . . . . . . . . . Distribution of Alaria p y l a i i Distribution of Alaria taeniata . . . . . . . . . . . . . . . . . . . . . . Distribution of Alaria tenuifol..a Illustrations of Alaria angusta . . . . . . . . . . . . . . . . . . . . . . Illustrations of Alaria crassifolia Illustrations of Mario cr-ispa — ......... Illustration^.;df Alaria esculenta Illustrations of Alaria fistulosa Illustrations of. Alari-. grandifolia Illustrations of Alaria marginata . . . . . . . . . . . . . . . . . . . . Illustrations of Alaria nana . . . . . . . . . . . . . . . . . . . . . . . . . Illustrations of Alaria ocbotensis . . . . . . . . . . . . . . . . . . . Illustrations of M a r i a paradisea . . . . . . . . . . . . . . . . . . . . . Illustrations of Alaria praelonga Illustrations of Alaria p y l a i i Illustrations of M a r i a taeniata . . . . . . . . . . . . . . . . . . . . . Illustrations of M a r i a tenuifolia The type specimen of M a r i a angusta . . . . . . . . . . . . . . . . . . The type specimen of M a r i a crassifolia The type specimen of M.aria crispa The type specimen of M a r i a esculenta The type specimen of M a r i a f istulosa The type specimen of M a r i a grandifolia The type specimen of Alaria marainata The type illustration of Alaria nana. The type specimen of M a r i a ochotensis The type specimen of M a r i a paradisea The type specimen of M a r i a praelonga The type specimen of M a r i a p y l a i i The type specimen of M a r i a taeniata . . . . . . . . . . . . . . . . . The type specimen of Alaria tenuifolia . . . . . . . . . . . . . . . Illustrations of type specimens of synonyms Illustrations of type specimens of synonyms continued :  193 195 197 199  201  203 204  205 206 207 208 209 210 211 212 213 214 215 216 218 218 219 219 220 220 221 222 221 223 224  225 226 226 228 229  vii ACKKOWLEDGEMEHTS I wish sincerely to acknowledge the help of Dr. R. F Scagel, of the University of British Columbia, who gave general direction to this study; of Mr. H. T. Powell, Marine Station, Millport, and Professor Y. Yamada, Faculty of Science, University of Hokkaido, who supervised and advised me during my work in Scotland and Japan respectively.  I wish to thank Dr. E .  Baardseth, Trondheim; Mr. G. Boalch, Plymouth; Professor H. Brattstrom, Espergrend; Mr. S. V. Hallsson, Reykjavik; Mrs. J . Jones (Kain), Port Erin; Dr. L. M. Lauzier, St. Andrews, New Brunswick; Professor T. Masaki, Hakodate; Miss 1. Munda, Trondheim; Professor Y. Nakamura, Muroran; Dr. M. Neushul, University of California, Santa Barbara; Dr. M. Parke, Plymouth; Cand. Real. P. Svendsen, Espergrend; and Dr. A. R. A. Taylor, Fredericton, for sending me field data and collections.  I wish to express my appreciation  to Mrs. Y. Butler, British Museum, for help in searching the literature there; to Dr. P. Dixon, University of Liverpool, and Mr. Ross, British Museum, for advice on nomenclatural problems; to Miss  Berners, Oslo,  for translating Swedish and Danish literature; to Dr. S. Nash, UBG Department of Mathematics, for advice on statistical problems; and to  Dr. E.  Froese, UBG Computing Centre, for help in writing programs. During my work.on this problem, I have enjoyed the greatest hospitality from many people in many places.  In particular, I wish tc thank for their  assistance the staff of the various herbaria I visited:  the University of  California, Berkeley (UC); Botanical Museum and Herbarium, Copenhagen ( G ) ; School of Botany, Trinity College, Dublin (TCD); the Royal Botanic  Garden,  Edinburgh (ED); Department of Botany, University of Glasgow (GL); Goteborgs  viii Botanisk Tradgard (GB); Farlow Library and Herbarium of Gryptogamic Botany, Harvard University (FH); Royal Botanic Gardens., Kew (K); Botanical Museum and Herbarium, Copenhagen (c); Rijksherbarium, Leiden (L); the Hartley Botanical Laboratories, University of Liverpool (LXVU); the British Museum (Natural History), London (BM); the Linnaean Society of London (HOT); the Botanical Museum and Herbarium, Lund (ID) [Agardhj ; the Fielding Herbarium, Druce Herbarium, Department of Botany, Oxford (QXF); the Laboratoire de Cryptogamie, Museum National d'Histoire Naturelle, Paris (PC); the Plymouth Laboratory, Plymouth [Ply]; Botany Division, Natturugripasafn rikisins, Reykjavik (ICEL); the Department of Botany, University of St. Andrews (STA); the Faculty of Agriculture JFac. A g . , Sap.], and the Botanical Institute, Faculty of Science, University of Hokkaido, Sapporo (SAP); the Botanical Institute, Faculty of Science, University of Tokyo (Ti); Botanical Department, Naturhistoriska Riksmuseum, Stockholm (S); Botanical Department, Museum of the Royal Norwegian Society for Science and Letters, Trondheim (TRH); V&xtbiologiska Institutionen,  University of Uppsala (UPSV). Material  was sent to me on loan from many of these, and also from the Department of Botany, University of Minnesota (MIN) and the New York Botanical Garden (NY), which I was unable to v i s i t . their personal herbaria;  I should like to thank those who showed me  Dr. S.' Lund, Copenhagen; Dr. Th. Christensen,  Copenhagen; Mr. E . Jaasund, Gj^teborg; Dr. R. T. Wilce, Amherst, Massachusetts; and Dr. M. S. Doty, Honolulu.  In addition, I should like to express  my most sincere thanks to the innumerable people who gave me help and hospit a l i t y during my field work. Finally, I wish to acknowledge the financial assistance of a National Research Council of Canada Studentship during the years 1959-60 and 1961=63;  ix and a British Council Scholarship, I96O-6I.  In addition, financial help with  field expenses in British Columbia, Alaska and Iceland was provided by the Institute of Oceanography, the Department of Botany of the University of British Columbia, and the Defence Research Board of Canada (DRB 9520-14).  _ 1 _  INTRODUCTION Scope of the problem  ,  Several species of the genus Alaria grow in British Columbia (Scagel, 1957, P- 108).  The writer found these species difficult to recognize in the  field and confusing to identify from the literature,,  Further study of the  literature indicated that uncertainty and, as a result, controversy concerning Alaria were widespread in the past and in areas other than British Columbia. The genus Alaria had long enjoyed the reputation of being a difficult subject, for taxonomic study (Ruprecht, 1851,.p. 35^; Yendo, 19.19, p. 1.; Setchell and Gardner, 1925v P» 6 3 6 ) . The present study of the genus was undertaken, on the assumption that this difficulty had a real basis in systematics. There appeared to be a number of f a i r l y distinct taxa within the genus.  It  also appeared probable that these taxa have a complicated interrelationship. A single variable taxon^, or a number of completely discrete taxa, would not cause the difficulty which appeared to exist.  The purpose of the present  investigation was to study the variation, and the geographical distribution of the various taxa within the genus Alaria. and from this study to suggest how these taxa might have evolved. Several practical features make Alaria a suitable plant for detailed taxonomic study.  It is large enough, to have a number of macroscopic features  by which taxa can be recognized in the f i e l d .  It is small enough usually to  be preserved as an entire specimen on a herbarium sheet.  Its morphological  complexity, relative for instance to that of laminar i a . besides providing a number of features for taxonomic study, implies a higher degree of organization in the plant.  The stability of this organization suggests that the plant has  2 a certain degree of independence from the effects of i t s environment.  A cer-  tain such degree of independence is necessary for the genus to be appropriate for a study of the effects of environment on taxonomic c r i t e r i a .  The compli-  cated interrelationships between the taxa within Alaria make i t a useful subject for a study of evolutionary processes^ even though such a situation requires an unusually extensive treatment of nomenclature. Historical background The species later designated (Silva, 1952, p. 259) as the type species of Alaria. Fueus esculentus. was described by Linnaeus from the Norwegian coast (17^7, p. 135). Bory de Saint Vincent (1826) and Bachelot de l a Pylaie (1829) described a number of species from material brought back from the Newfoundland area by De l a Pylaie. a source of confusion ever since.  These species have been  Postels and Ruprecht (l84o) described  several species from the North Pacific.  J , G. Agardh (1848) was the f i r s t  to attempt a systematic treatment of the genus.  Ruprecht (1851)des-  cribed species from the Okhotsk Sea area and also discussed the systematica of the genus. bergen.  In 1872 Agardh described some new species from Spitz-  Kjellman (1883a, 1883b, 1889) added a large number of species,  mostly from Arctic and Pacific Russia.  Most of the material he used for  these'descriptions was collected during the voyabe of the "Vega" through the Northeast Passage.  Saunders (1901a, 1901b) and Setchell and Gardner  (1903) described more species from Pacific North America. named a number of species from northern Japan. the genus Alaria.  Miyabe (1902)  Yendo (1919) monographed  His coverage of the literature and of herbarium ma-  t e r i a l , and his wide field experience were exceptional for his time.  ^  However, the key he presented is not very workable since the qualitative characters used are not precisely defined.  The descriptions given of each  species are often difficult to compare since different features are described in different cases.  Setchell and Gardner (1925) discussed the genus  again, giving a completely unworkable key which consists almost entirely of quantitative characters expressed in non-nuaerical form (for instance, 'broad', 'long', and 'wide').  Taxonomic work with the genus Alaria has  been quiescent since Setchell and Gardner's time, except for the description of the genus Pleuropterum by Miyabe and Nagai (1932), a genus reduced to synonymy under ALarja below (p. hi).  A number of new synonymies and re-  combinations, or rejection of the synonymies and recombinations of preceding workers, in such general works as those by Nagai (19^0), Taylor (1957)> and A. D. Zinova (1953)> indicate that uncertainty s t i l l exists in the taxonomy of the genus. Almost a l l research on the genus Alaria. has consisted of taxonomic studies.  For the rest, a complete study of the anatomy of the plant was  made by Wille (1897) and further investigations into some aspects of anatomy were performed by Baardseth (1956), and Hishibayashi and Inoh (1963). The l i f e cycle was studied by Sauvageau (1916) and by Kanda (1936). dealt briefly with growth patterns.  Yabu (1957) treated' the  F a l l i s (1916)  cytogenetics.  No research appears to have been undertaken concerning the physiology of the plant.  Experimental ecology has been covered in some aspects by Sundene  (1962). Approach followed in this stud: Since there is an arbitrary element in many aspects of taxonomy, i t is necessary to outline the basic approach followed in this study.  - 4 -  As much of the past confusion has derived from a nonchalant attitude to nomenclature, the rules of the International Code of Botanical Nomenclature (Lanjouw, 1961, abbreviated after this as "The Code") have been followed s t r i c t l y in this work.  At the same time, an effort has been made to interpret  the rules where possible to avoid disadvantageous changes in names currently used.  It appears that during the past many taxa have been founded on l i t t l e  or poor material, with the result that the description is insufficient for practical use.  In the present work, such taxa are listed but are not given  systematic consideration in order to avoid the confusion which would arise from the use of a large number of taxa of doubtful status. Names with useable types and descriptions were considered from a systematic point of view.  It was evident from a preliminary study that any popu-  lation of Alaria contains a few individuals which appear to be aberrant in their form, and which thus simulate representatives of some other population. In view of this, no species which is represented by a single specimen, or even a single collection, was given serious consideration.  A l l species con-  sidered in the systematic study have been collected either from more than one locality, or from one locality more than once. Morphology has provided most of the criteria used in past descriptions and is also the basis of the systematic treatment of the genus presented here.  Morphological features are the only ones which can be taken from  habit illustrations, which can be observed in herbarium material without damaging i t , and which can be observed in the field in a large number of i n d i viduals during a short space of time.  Morphological characters alone are not  necessarily the best for taxonomic purposes, but they are the most suitable for a preliminary study.  Anatomical, ecological and physiological investi-  gations, which usually give a large amount of information about a few  individuals, are probably more suitable for later, detailed studies. Finally, i t was considered that the entities delimited should usually have a meaningful geographical distribution and relationship to environment. Disjunctive  geographical-.! distributions, and apparently i l l o g i c a l environ-  mental relationships, were examined with particular care.  Such distributions  and relationships might sometimes represent errors in the interpretation of systematics Instead of a real situation. Terminology A generalized illustration of an Alaria plant is shown in Figure 1. At the distal end is a lamina (A), consisting of midrib (F) and blade (E). At the proximal end of the blade is the transition zone (G), which is the site of the intercalary meristem.  Proximal to the transition zone is the trunk.  This is a new term, introduced here because the word "stipe" has been used ambiguously in the past, either as a synonym of "trunk" here, or in the same way as "stipe" is used below.  The upper part of the trunk, bearing the sporO'  phylls or showing obvious scars of sporophylls which have fallen off, is the rachis (B).  The sporophyll (H) is attached to the rachis by a petiole (K).  A reproductive sporophyll consists usually of a sterile area, and a fertile sorus (J). The remaining part of the trunk is the stipe (c). The organ of attachment of the stipe to the substrate is the holdfast (D), consisting in most cases of branched haptera (L). Definition of measurements For s t a t i s t i c a l purposes, the morphological features used as diagnostic characteristics (Appendix I (2)) were reduced to a number of quantitative measurements.  These measurements (the numbers of which correspond to those  in Figure l ) were made as follows: 1.  The width of the midrib was measured immediately distal to the point  where the blade ceases rapid increase in width.  This is a well defined posi-  tion except in the case of Alaria fistulosa. 2.  The width of the rachis was measured at its widest point, in the  same plane as that of the lamina. 3.  The width of the stipe was midway along its length (measurement No.  5) in the plane of the lamina. 4.  The length of the rachis was measured from the narrowest point of  the transition zone to the sporophyll scar furthest from i t . 5.  The length of the stipe was measured from the proximal end of the  rachis.to the hapteron nearest i t .  This measurement is doubtful in old  plants and with Alaria ochotensis. because of difficulty in deciding which projections are actually scars. 6. The number of sporophylls was counted to include definite stumps and scars, but not to include both lobes of definite bifurcations of the sporophyll.  This count is doubtful in old plants with many scars, but doubtful  decisions were kept consistent with similar decisions in making measurement No. 57.  The width of the sporophyll petiole was measured at its narrowest  point in the plane of the sporophyll.  The sporophyll with the greatest area  was used. 8.  The width of the same sporophyll as that used in No. 7 &s measured w  at its widest part, at right anglesLto/.the .major .axis . This measurement is occasionally doubtful with Alaria crispa. in which the sporophylls are sometimes nearly cylindrical in cross-section. 9. The length of the same sporophyll was measured from its point of attachment to the rachis, making allowances for missing pieces of the sporophyll which obviously had.been recently.detached.  10.  The angle of the base of the lamina was- subtended from the widest  part of the blade to the most distal sporophyll. The width of the origin of the lamina was measured at the narrowest  11.  part of the transition zone, in the plane of the lamina. 12.  The depth of the stipe was measured at the same place as No. J,,  in a plane perpendicular to that of the lamina. 13.  The number of fertile sporophylls included every sporophyll which  showed a sorus as a dark patch when held up against the light. Measurements No. 6 and 13 have no units. dians .  No. 10 was measured in r a -  The remaining measurements were made in centimeters  In the statis-  t i c a l analysis, unity was added to No. 6 and 13, to prevent the data equalling zero. Pattern of growth A study of the morphology of Alaria is of l i t t l e value without an understanding of the manner in which the plant grows.  The growth in one of the  plants studied in the field at Millport, Scotland, is shown in Figure 2. As w i l l be discussed below (p. 8 ) , the details of the growth patterns are not consistent throughout the genus.  The pattern here described is that of the  type species, Alaria esculenta. Most of the active elongation occurs in the transition zone, from which new tissue is formed distally towards the stipe and proximally towards the trunk.  Some elongation also occurs in the lower midrib, but distally this  growth gradually diminishes to nothing.  The gradient of growth rate is much  steeper on the proximal side of the transition zone. found proximal to the distal sporophylls.  No elongation was  Tissue is lost from the distal  end of the lamina, sometimes continuously as a result of decay or erosion, sometimes discontinuously through loss of fragments.  Decay or erosion occurs  - 8 more rapidly during periods of relatively high sea temperatures (approaching l6°C).  The greatest growth occurs during late spring and early summer, but  some elongation occurs throughout the year.  These variations in growth rate  during the year were the cause of controversy about Alaria "shedding its lamina" in spring (Boergesen, 1907).  The pattern of intercalary growth per-  mits the age of the proximal region of the individual to be much greater than the age of any tissue above the transition zone.  Thus radical changes in the  form of the lamina may occur during the growth of the plant.  The blade is  produced obliquely from the midrib, but there is also some diffuse growth of the blade i t s e l f in the same direction. sion at the lateral margins of the blade. growth throughout their length.  Tissue is lost.through decay or eroThe sporophylls have a diffuse  This growth is determinate, unlike that of  the lamina, so that the maximal size of the sporophyll is achieved relatively quickly and is not the resultant of proximal growth and distal erosion. After the spores have been released, the sporophylls may die back, but more often they f a l l off entire. These patterns of growth are not consistent for the whole genus.  The  rates of growth of the trunk and lamina were studied in Alaria esculenta at Millport, Scotland, A. tenuifolia at Brockton Point, near Vancouver, B.C.", &. marginata at Muir Point, near Sooke, B . C . ; and 4- nana at Botany Beach, near Port Renfrew, B.C. The data for the last three species are shown in Figure 3. A limited amount of replication at Millport, Brockton Point, and Muir Point, together with additional observations of A. marginata and A. nana at Glacier Point, B . C . , indicate that the patterns shown in Figure 3 occur consistently in neighbouring populations of the same species.  However, these replications  are too close geographically to the original observations to prove that the patterns shown are general in the relevant species throughout their entire  _ 9distributions. Other differences in growth between the species were discovered.  Cer-  tain characteristics of A. esculenta (growth in the proximal part of the midrib,  production of the blade from the midrib, and diffuse growth in the blade)  could not be detected in the other three species studied.  Undoubtedly  growth exists within a few centimeters of the transition zone in a l l species, but i t must be in a very restricted area with some.  Studies of mature sporo-  phytes maintained in laboratory culture showed that there can be a large amount of diffuse growth in the rachis of A. nana.  More than a half of the  elongation of the* trunk occurs between the proximal and distal sporophylls. The data in Figure 3 support the generalization that Alaria plants have a longevity of one year or more, since the growth rates require about a year to result in plants of the size found in the f i e l d .  Thus i t is a reasonable  assumption that collections made in the same geographical locality and vert i c a l position belong to the same population.  Any differences found between  them may be attributed to developmental or environmental effects. Distribution and ecology of the genus The overall distribution of the genus Alaria is indicated in Figure 4. Alaria is associated with the cold waters of the Northern Hemisphere. There are no evident northern limits  The southern limits of the genus are close  to the 20°C isotherm of maximum sea temperature in California, New England, Brittany (France), and north-eastern Honshu (Province Mutsu, Japan). One group of specimens (L910 153-1677, -I685, -1740) is labelled as coming from South America.  These are only fragments of lamina, so positive identification  is impossible.  They are identified as Alaria fistulosa. which they certainly  are not.  The last of the three listed is probably a Pleurophycus.  It is  reasonable to assume that some error has occurred in the locality also.  10  The relationships between any marine intertidal organism and factors of its.environment may be exceedingly complex. These factors have been divided functionally into 'presence-and-absence factors' and 'causal factors' (Chapman, 1 9 ^ 2 , p. 2 ^ 3 ) .  Presence-and-absence factors are those which permit or  prevent an organism from becoming established in a particular locality.  Three  possible presence-and-absence factors (sandy shores, ocean currents, and seaice), which may play a role in the distribution of Alaria. are shown in Figure 5 - These presence-and-absence factors are presented because of their suggested role in the increase or restriction of gene flow between populations and their possible effect on the evolution of species (pp. 80 f f . ) . Sea temperature and salinity are the causal factors which are most commonly discussed (Scagel,  1 9 6 3 ) .  They are the oceanographic data which are  most commonly available, and have an excellent descriptive value, in that water types and water masses are defined by them.  They may or may not be in  themselves determining factors i n algal distribution. The sea temperature and salinity at the four southern limits of Alaria. in California, New England, the English Channel, and Japan, are presented in Figures 6 , 7 , 8 , and 9 . As far as possible these data have been made comi parable.  In each figure, values are given for a locality close to the extreme  southern limit where Alaria actually does grow (A), from a locality somewhat more sheltered from surf where Alaria does not grow (B), and from an exposed locality south of the extreme limit (C).  Variations are shown over a period  of two years, which is approximately the l i f e span of an Alaria plant.  As  far as possible, the data have been taken from the surface near the beach, and represent oceanographic conditions actually experienced by the plants there.  The localities of A, B, and C are shown in Figures 1 9 , 2 1 , and 2k,  on the distribution maps of the species involved: Alaria crassifolia. A.  - 11 esculenta. and A, marginata.  Data for B and C are not available from Japan.  Sundene (1962) has shown from both distributional evidence and experiments in the field and laboratory, that a sea temperature of l6°C is fatal to mature sporophytes of the species with which he worked (A. esculenta in Norway).  A consideration of Figure 9 shows that Sundene's results do not  apply to Alaria in general, since a Japanese species (A. crassifolia) is associated with summer sea temperatures of up to 2k°C,  In A. crassifolia. the  plants become f e r t i l e only during the winter, when the sea is colder. A similar adaptation appears in A. tenuifolia. which is fertile only during the winter in sheltered l o c a l i t i e s , such as Brockton Point, where the summer sea temperatures may reach lS^C. In the more sheltered parts of Juan de Fuca Strait and Johnstone, Strait, B. C , where the:, sea temperatures, never exceed l 6 ° C . A. ..tenuifolia is f e r t i l e during the summer.  In areas where the  sea temperature never rises above 10°C, Alaria is usually fertile in the later summer, when the sea is warmest, but since these areas have seldom been visited at any other time, i t is not possible to be sure that Alaria is f e r t i l e only during the summer. Field observations in the temperate region show that most species of Alaria have an upper limit in the lower intertidal zone, and a lower limit somewhere in the upper subtidal zone.  Typically, the plant is found on  coasts having a moderate to severe exposure to surf. surf, the higher Alaria grows in the intertidal zone.  The more severe the The present study does  not cover the subtidal populations in detail, but the deepest records in various l o c a l i t i e s , as noted in the literature, and the lower limits, as communicated to the writer, are shown in Figure 10.  A trend is apparent,  in that the lower limit tends to become deeper towards the north.  In the  Arctic, Alaria is usually absent from the intertidal zone, according to  ~ 12 -  Wilce  and Svendsen (pers. comm.). Certain species occupy restricted habitats.  Alaria crispa and A. nana  occupy belts in the mid-intertidal zone which are discontinuous with the belt of other species of Alaria which occur close to low water mark. A. crisna is found in the Aleutian Islands, which form an area with an exceptionally high degree of cloud cover.  A. nana is found in western North Amer-  ica in localities with extreme exposure to surf.  Thus i t appears that the oc-  curence of both species is associated with conditions of high humidity. A. tenuifolia is found close to low water mark in very sheltered areas, but only in localities subject to intense water movements because of t i d a l currents.  A. fistulosa is attached in the subtidal zone and has buoyant cham-  bers in the midrib so that the distal portion of the lamina floats on the surface of the water. Historical changes in distribution Available herbarium material collected during the past two centuries can be used with validity to define geographical distributions only i f i t is reasonably certain that the distributions of the species have not changed during this period of time.  Unfortunately the policy of using only extant  specimens makes the extent of distributions in the past difficult to check. A distribution consists essentially of both positive and negative records. A negative record (the information that the plant does not grow in a p a r t i cular locality) can exist only in the literature, and usually is not explic i t l y given even there. In two situations i t is possible to use information in the literature about past distributions.  Alaria fistulosa is a very distinctive and ob-  vious species, and can scarcely be misidentified or overlooked where i t  - 13 occurs.  The distribution of A. fistulosa in southern Alaska given by  Saunders (1901b, p. 426) and Frye (1915> p. 67) can be compared with that of the present,  Alaria esculenta is the only species of.Alaria in southwest  England, so there is no difficulty in identifying i t .  Alaria is close to  its extreme southern limit here, and the area has been well covered by workers from the Plymouth Laboratory, so i t is reasonable to expect that the plant has been found and noted where i t occurs.  The distribution given by  Batters (1902, p. 48) can be compared with that.of the present. Saunders reported drift Alaria fistulosa in Southern Alaska only from Wrangell, Juneau, Yakutat Bay, Prince William Sound, and Cook Inlet.  The  only locality in southeast Alaska where he found the species attached was Glacier Bay, and he collected specimens from there (NY). Frye (1915) reported a great number of kelp beds composed in whole or in part of A. f i s t u losa throughout southeast Alaska (Figure l l ) .  In i960 the writer found  A,, fistulosa growing as far east as the Kenai Peninsula in the Gulf of Alaska. It was absent from the v i c i n i t y of Cordova, Yakutat, Juneau, Sitka, and Ketchikan.  It was probably absent from localities between these, which  were observed from the air in clear weather at relatively low altitude (less than 2,000 feet).  The writer's findings in i960 are consistent with Saun-  ders' in 1899 (Saunders, 1901b, p. 426). While the writer has been unable to find any herbarium specimens which appear to be relevant to Frye's records, i t does not seem possible that his divergent account is without foundation.  Frye (1915) presented photographs of some of the kelp beds, of which  Plate XXXIII of Keku Inlet, definitely shows A. fistulosa. Crandall (1915) show more pictures of the plant.  Frye, Rigg, and  It appears that there have  been quite large, relatively short-term fluctuations in the distribution of A. fistulosa.  It is impossible at this time to suggest what factor or  - 14 -  factors were responsible for these fluctuations. The distribution of Alaria esculenta in southwest England is shown in Figure 12„  Alaria has disappeared from the easterly parts of its former  area of distribution. better known.  In this instance, the environmental factors are  Southward ( i 9 6 0 ) found an increase in mean sea temperature  of about 0 . 5 C°in the western English Channel during the previous 50 years. Southward and Crisp (.1954) had already suggested a rise in sea temperatures as a cause of the retreat of certain typically northern intertidal animals from the easterly part of their area of distribution in the English Channel. The fact that Sundene ( 1 9 6 2 , F i g . 8 ) ignored these historical changes may have caused some minor Inaccuracies in his picture of Alaria distribution. It is difficult to compare the data used, and the conclusions reached, by Sundene and by Southward and Crisp.  The former discussed extreme values of  sea temperature while the latter used long term means. These two changes in distributional area are of a f a i r l y small scale relative to the total area occupied by the species involved.  It seems  unjustifiable to reject the distributional evidence afforded by past c o l - . lections.  However, added caution is indicated i f an attempt is made to re-  late past distributions to factors in the present environment.  TAXONOMY General Taxonomy, the science of the classification of organisms, has been divided into nomenclature and systematics.  Nomenclature is the study of the  names given to the categories under which organisms are classified.  Systema-  tics is the study.of the organisation and evolution of the categories proposed . Nomenclature in the present study began with the preparation of a l i s t of a l l the names of taxa which have at any time been included under the genus Alaria or its synonyms. below in four tables.  For convenience these names are presented in the text Names for which the type material found is considered  adequate for systematic use are listed in Table 1, names with inadequate type material in Table 2, names for which no type material could be found in Table 3. Names which are illegitimate or removed from the genus are listed in Table k. In view of the criteria which are used in the present study, adequate type material is generally considered to be either a specimen, fert i l e and complete except for the distal part of the lamina, or a photograph of such a specimen.  A l l the names in the f i r s t three tables appear to be l e -  gitimate, but only those in Table 1 can be thoroughly evaluated from a point of view of systematics.  Names in Tables 2 and 3, which lack descriptions and  type material sufficient to permit an examination of their relation to other taxa, must remain doubtful for the present. betical by specific epithet.  The l i s t i n g i n the tables is alpha-  Where subspecific names are involved> these are  listed after the prior combination i f these were new recombinations, but after the specific epithet i f the subspecies were new descriptions dividing off a part of the original species.  The effect of this arrangement is to  -  16  place nearly a l l taxa which have the same nomenclatural type adjacent t o one another on the l i s t . Notes on the t y p i f i c a t i o n of the species treated i n d e t a i l (pp. 4 5 f f . ) are made below where these species are treated i n the t e x t .  Notes on the  t y p i f i c a t i o n of a l l other taxa and, where appropriate, notes on the adequacy of the type and the legitimacy of the name are presented i n Appendix I.  Type l o c a l i t i e s are shown i n Figure 4 .  standard abbreviations (Lanjouw and Stafleu,  Herbaria are referred to by the 1956).  Two herbaria not  l i s t e d there are abbreviated as follows: Faculty of Agriculture, U n i v e r s i t y of Hokkaido, Sapporo (Fac. Ag., Sap.) and the Plymouth Laboratory ( P l y ) . The type method was not used extensively with regard to A l a r i a before Yendo's monograph  (1919).  In e a r l i e r cases where there i s no ambiguity  concerning the intention of the o r i g i n a l author, because one element only i s involved, "holotype", "type species", and "type l o c a l i t y " are used i n the discussion even though i n the nature of things the designation was not made e x p l i c i t l y by the o r i g i n a l author (Lanjouw, 1 9 6 1 , p. 6 4 ) . The treatment of the systematics of the genus was l a r g e l y independent of the study of the nomenclature.  The c r i t e r i a used i n describing the names  (Appendix I ( 2 ) ) were reduced so f a r as possible to standard measurements (p. 5 ) . These measurements were analyzed s t a t i s t i c a l l y (pp. 3 0 f f . ) .  The  s t a t i s t i c a l analysis, f i e l d observations, and examination of herbarium material indicated the existence of fourteen species i n the material a v a i l able f o r study.  These fourteen species appear i n a key (p. 4 3 ) and each i s  treated i n d e t a i l . The d e t a i l e d treatment i s , as f a r as possible, consistent so that the data on each species are comparable to the data on every other.  These data  include the synonyms, notes on the type, geographical d i s t r i b u t i o n , r e l a t i o n  - 17 to other species and environment, and representative specimens examined. The synonyms of the fourteen species treated included a l l the other taxa in Table 1 (p. 19).  The type localities of the species and its synonyms, l o -  calities of "representative specimens examined", and localities of samples used in s t a t i s t i c a l analysis (Appendix II, Table I) substantiate the geographical distribution given for the species.  Additional l o c a l i t i e s , a l l  based on preserved or herbarium specimens, are given in Appendix III, Table 1. The figures accompanying the treatment of each species include an i l lustration of the type specimen together with the trunk and a representative sporophyll of a specimen intermediate to each of two similar species, and a map of the geographical distribution.  Illustrations of the type spe-  cimens of the synonyms are grouped together in Figures 60 and 6 l . The geographical distribution of a species is the principal aspect of its biology discussed in the conclusions following, so particular attention has been given to avoiding mistaken, or at least inconsistent, locality records .  Only those localities where a relevant specimen has been examined  by the writer himself are cited.  In many cases this resulted in the ex-  clusion of many obviously correct localities noted in the literature, but i t was thought best te make no exceptions to this rule. The latitude and longitude of localities are given in Appendix III, Tables 1 and 2.  The localities are classified by geographical region.  These geographical regions are shown in Figure 4. Some of the more important places mentioned are entered on the maps which show the area in most det a i l : Hew England, Figure 21; Southwest England, Figure 12; Northwest Europe, Figure 21; Northern Japan, Figure 19; Saghalien, Figure 26; Kurile Islands, Figure 27; Southeast Alaska, Figure 11; British Columbia, Figures 24, 25* and 31; California, Figure 24.  The specimens included in the samples used for s t a t i s t i c a l analysis (Appendix II, Table l ) are deposited in the Phycological Herbarium of the University of British Columbia (UBC).  Many of the new records listed  (Appendix III, .Table l ) from British Columbia and Alaska are based on smaller collections made by the writer and also deposited at UBC.  - 19 List o£. names discussed Table 1. - Names for which adequate type material was found. No. Generic Name. Epithet. Reference. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41  Type specimen  Alaria angusta Kjellm., I889, p. 38. Alaria corrugata Miyabe, 1902, p. 55Alaria crassifolia Kjellm., in Kiellm. et Petersen. 1887, p. 276. Alaria crista Kjellm., 1889, p. 37, AlarpLa curtipes Saunders, 1901a, p, 5 6 l . Agarum delisei Bory, 1826, p. 1§4. Alaria delisei (Bory) Grev., I83O, P- xxxix, Orgya delisei (Bory) Trev., 1845, P. 28. Alaria aolichorachis Kjellm., 1883a, p. 217. Alaria dolichorachis f. longipes Miyabe, 1928, p. 957. .Fucus esculentus L., 1767, p. 135. Ceramium esculentum (L.) Stackh., 1797, P. 24. Musaefolia esculenta (L.) Stackh., 1809, p. 53Orgyia esculenta (L.) Stackh., l 8 l 6 , p. 8. Laminaria esculenta (L.) C. Ag., 1817, p. 13. Phasgonon esculentum (L.) Gray, 1821, p. 385Agarum esculentum (L.) Bory, 1826, p. 194. Podopteris esculenta (L.) De l a Pyl.. 1829, P. 23. Alaria esculenta (L.) Grev., I83O, p. 25. Orgya esculenta (L.) ^., Trev., . = v., 1845, p. 28. Alaria esculenta f . fasciculata Stroemfelt, 1886, Po 38. Fucus esculentus var.^ minor Turner, l 8 0 2 ]3. 104. Phasgpnpn. esculeRtUffl/sminus Gray, 1821, p. 385Pleuropterum. fasciculatum Yamada, 1935, P- 16. Pleuropterum paradiseum f. fasciculatum (Yamada) Miyabe §_£. Nagai, 1933, P- 100. Alarja fjstulpsa P. §i.R., 1840, p. 11. Orgya fistulosa (P. et R.) Trev., 1845, p. 28. Phasganon fistulpsum (P. et R.) Rupr., 1851, p. 355. Alaria fistulosa f. platyphylla Setchell i n Collins, Holden. et Setchell, 1901, XLII. Alaria f r a g i l i s Saunders, 1901b, p. 425. Alaria f r a g i l i s f. bullata Saunders, 1901b, p. 425. Alaria grandifolia J . Ag., 1872, p. 26. Alaria p y l a i i var. grandifolia . ( J . Ag.) Jonnson,, • 1904, p. 20. Alaria lanceolata Kjellm., 1889, p. 39. Alaria latlcosta Kjellm., 1889, p. 40. Alaria linearis Stroemfelt, 1886, p. 38. Phasganon macropterum Rupr., 1851, p. 353• Alaria macropt'era (Rupr.) Yendo, I919, p. 79. Alaria marginata P. ejt R., 1840, p. 11. Orgya marginata (P. e£R.) Trev., 1845, p. 28. Phasganon marginatum (P. e_fc.R.) Rupr.,1851, p. 355.  UPSV Fac. Ag., Sap. UPSV LD 2081 FH PC 11 11  LD 2251 Fac. Ag., Sap, LINN 1274.63  AJ  8  S K  11  SAP 15356 S  11  uc 96594 FH 11  LD 2256 LD 2083 UPSV S LD 2085 11  S  11  - 20 —  Table 1 (Cont.)  Ha*. 42.  43.  44. 45. 46. 4748. 49. 50. 51. 52. 53. 54. 55. 56. 57-. 58. 59. 60. 61.  62. 63.  64. 65. 66.  Generic Name EpliheJ^_geference,  Type specimen  l9  Laminaria musaefolia De l a Pyl.,, 1829, p 31. A l a r i a musaefolia (De l a Pyl.) J . Ag , 1872, p. 2 3 . Maria, esculenta f . musaefolia (De l a Pyl.) Kjellm., 1883a, p. 2 1 2 . Laminaria musaefolia @ remotifolia De l a P y l . , 1829 P. 3 5 . A l a r i a nana Schrader, 1903, p. 157. A l a r i a praelonga f . nana (Schrader) S e t c h e l l i n C o l l i n s , Holden §i S e t c h e l l , 1905, Wo. 1292. A l a r i a marginata f . nana (Schrader) C o l l i n s , 1913* p. 110. A l a r i a oblonga Kjellm., 1883a, p. 2 2 0 . A l a r i a ochotensis Yendo, 1919, p. 84. Pleuropterum paradiseum Miyabe et Nagai. 1932, P. 127 Pleuropterum paradiseum f . brevipes Miyabe et Nagai, 1933* P« 101. Orgyia pinnata (Gunn.) Ghobi, 1878a, p. 7 7 . A l a r i a esculenta f . pinnata (Gunn.) F o s l i e , 1886, p. 3 8 . A l a r i a praelonga KjelUn., 1889, p. 3 8 . Agarum p y l a i j Bory, 1826, p. 194. Iaminayia p y l a i i (Bory) De l a P y l . , 1829, p. 3 1 . Alaria pylaii (Bory) G r e V i , 1830, p. xxxix. Qrgya. p j d a i i (Bory) Trey., 1845, p. 2 8 . A l a r i a taeniata Kjellm., 1889, p„ 3 6 . S e t c h e l l , in, C o l l i n s , Holden, §i S e t c h e l l , X L V , 1901. Maria f . amplior, S e t c h e l l , i n S e t c h e l l g i Gardner, 1903* p. 2 7 4 . Fucus teres Good et Woodw., 1797, p. 140. A l a r i a v a l i d a Kjellm. et S e t c h e l l , i n S e t c h e l l et Gardner, 1903",~P- 2 7 8 . A l a r i a v a l i d a f . longjpes S e t c h e l l , i n S e t c h e l l et Gardner, 1903* p. 2 7 9 . A l a r i a vezoensis Miyabe, 1902, p. 54. ?  PC  11  0  PC Plate  24  LD 2084 TI Fac. Ag.,  Sap,  TRH LD 2077 PC 11  DPSV uc 96659 uc 96651 K uc 96663  UC 96666 Fac. Ag.,  Sap,  - 21 Table 2. - Karnes for which inadequate type material was found. No.  Generic Name. Epithet. Reference.  67.  Alaria cordata  Tilden, American Algae, 1898, No. 241.  68. Alaria e l l i p t i c a Kjellm., 1883a, p. 221. 69. Alaria esculenta f. australis Kjellm., 1883a, p.'202." 70. Laminaria esculenta var. Noltij Horn., 1837, P. 737. 71. Alaria esculenta H pinnatifida P.. g t R . , 1840, p. 11. 72. Alaria esculenta p l a t i f o l j a P. § £ , R . , 1840, p. '11. 73. Alaria flaaellaris Stroemfelt, 1886, p. 4 l . 74. Alaria macrophvlla Miyabe, 1902, p. 56. 75. Alaria membranacea J . A g . , 1872, p. 26. 76. Alaria p y l a i i a membranacea ( J . Ag.) Rosenvinge, 1893, P. 838. 77. Alaria platvrhiza Kjellm., 1906, p. 11.  Type  MIN 16748 LD 2082 Turner, 1802 Plate 117 C  FH LE S Plate 25 LD 2113 UPSV  Table 3. - Names for which no type material has been found. MJU  Generic Name. Epithet. Reference.  78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88.  Alaria esculenta^ angustifolia p„ g±.R., 1840, p. 11. Lamjtoarja, esculenfra. var. platvphylla De l a P y l . , 1824, p. 178. Laininflria,'escwlaJla. var. poiyphviia De l a P y l . , 1824, p, 179. Lamjnaria escuJLenta var. remotifolia De l a P y l . , 1824, p. 179. Laminarja esculenta var. taeniata De l a P y l . , 1824, p. 179. Fucus f i m h r l a t u s Gmelin, 1768, p. 200. Phasganon longipes Rupr., 1851, p. 353. Alaria marginata a musaeformis P. et R., 1840, p. 11. Laminaria musaefolia var. decipiens De l a P y l . , 1829, p. 36. Fucus polvphvllus Gmelin. I768. p. 206. Phasganon scoticum Walker l a N e i l l , 1815, p. 4.  - 22 -  Table 4. - Names which are illegitimate or removed from the genus.  flamsfipJAbsig  Ho*  fiSBerjc  89. 90. 91.  Pha,ggan.pn, ajatum Rupr., 1851, p. 353. Alaria amnlexicaulis Martens, 1886, p. 144. A l a r j ^ caulescent Savage, 194-5, p. 200.  92.  93. 94.  95. 96. 97. 98. 99100. 101.  102.  103. 104. 105. 106.. 107.  flefexence^.  .Aiao^ cflfliata • Faiiis, 1916, f i . ' 2 5 .  Laminarja despreauxjj Bory, i a Rupr.., 1851, p. 357Alaria despreauxii (Bory, 4a Rupr.) J . Ag., in Nordenskiold, 1872, p. 1081. Alaria dolichorachis f. typica Nagai, 1940, p.- 110. Alaria esculenta f. typica Kjellm., 1883a, p. 3^1. Alarja egculen,ta, var. l a t i f o l i a f. singuLaris Rupr. i n Yendo, 1919, p. 89. Alarja fistulosa f. stenpphylla Setchell, in Collins, Holden, gt Set- . chell, B, XLIII, 1907. Laminaria linearis De l a Pyl., 1829, p. 37. Laminaria megalopteris De l a Pyl., 1829, p. 29. Jm\mr1& n p l t i i (Horn.) J . Ag., in. Yendo, 1919, P. 109AlarjA p ^ n n R t ^ f i d a Harvey, i860, p. 329. Fucus^ pinnatus Gunn., I766, p. 96. A2ssi& Exlaii f . iyjaisa Rosenvinge, 1893, p. 838. Alaria striata J. Ag., i& Yendo, 1919, p. 93. Alaria tenuifolia f. iyjgica. Setchell, i& Setchell ei. Gardner, 1903, p. 272 Fucus tetragonus Good, et Woodw., 1797, P. l40.  - 23 -  The discriml.nant and distance functions Before the initiation of the present study, the writer had collected a number of individuals of Alaria. from the area around Juan de Fuca Strait, and had made some observations of the original populations in the f i e l d .  The  observations, together with examination of the individuals already collected, indicated that much apparently random variation exists in the genus. When a sufficient number of individuals are examined, individuals intermediate in form between any designated categories can nearly always be found.  Within  any large sample, an occasional, apparently divergent individual can usually be found.  Thus, many data contain l i t t l e information which can be used for  classification.  Large numbers of individuals in each sample, many samples,  and relatively sophisticated techniques are needed. Examination of a large number of individuals is necessary in a survey where the taxa involved do not correspond to a number of clearly defined populations, which can always be distinguished from one another.  There are  a number of possible techniques by which this examination may be accomplished. The reasons why a study of morphology was chosen for the survey, rather than anatomical, ecological, and physiological c r i t e r i a , are given on page 4. Morphology can be studied qualitatively or quantitatively. criteria have been most commonly used in the past.  Qualitative  They are on the whole  unsuitable in a genus such as Alaria. where one taxon seems often to merge into another, since they do not lend themselves to precise definition in such circumstances.  An example of this occurs in Yendo's key, where he  uses "sporophylls cuneate" as the opposite of "sporophylls long cuneate" (1919, P- 75) •  Qualitative data can be used in statistics by means of  scoring (Anderson, 194-9), but in such cases the apparent precision of the  numerical data used to some extent must be the product of subjective judgement on the part of the observer. expressed in numerical form.  Quantitative data are not necessarily  They are sometimes confusing in non-numerical  form as, for example, when Setchell and Gardner used "blade long and broad" as the opposite of "blade short and narrow" in their key (1925, P« 636). Quantitative data are much more easily converted into numerical form than are qualitative; indeed they usually are observed in that form directly, as measurements. In the present study, a large number of data were used.  Measurements  were made of 13 different features of each of approximately 2,000 i n d i v i duals.  These 2,000 individuals were chosen as representative of the ap-  proximately 5,000 collected, which in turn were chosen carefully in the field to represent the populations there. herbarium specimens were examined.  In addition, approximately 2,500  Statistical techniques were made even  more necessary by the fact that variation i t s e l f was a subject for study. The discriminant and distance function analysis is the standard statist t i c a l technique for use in a classification problem, such as the present, which Involves the simultaneous use of a number of related characteristics. It was decided to use i t here, as i t also included the calculation of the covariance matrices needed for a study of variation.  The discriminant and  distance functions were originally proposed by Fisher (193&) and Mahalanobis (1936).  Recent discussions and examples of their use can be found in  Lubischew (1962) and Elaekith (1957). A standard text on the subject is by Rao (1952, p. 286). A discriminant function is a means of pooling the discriminating power of a number of quantitative characteristics, when distinguishing between  - 25 -  certain populations similar to one another, where no single characteristic is sufficiently non-overlapping between populations for the distinction to be made with the required degree of confidence„  A distance function des-  cribes the relationships between these populations in terms of vectors between them in an n-dimensional hyperspace, where n is the number of the characteristics used. The terms "discriminant" and "distance function" have a somewhat imprecise < usage" in the biological  literature, as they describe groups of  techniques used for a certain purpose, and not unique sets of calculations. They are often parts of the same program as in the present case and so are referred to here as "a discriminant and distance function analysis". Techniques used in this study were based on the computer program written and placed i n the library of the University of British Columbia Computing Centre by Dr. D. F . Snyder.  This program i t s e l f was based upon the methods  of calculation developed by Nash and Jolicoeur (unpublished ms.). If the data from the different populations be considered as plotted against one another in n,-dimensional hyperspace, one would expect them to appear grouped by populations as ellipsoids with major and minor axes oriented in the same directions and having the same proportions to one another.  In more mathematical language, these,populations are assumed to  have the same dispersion matrix.  This dispersion matrix is estimated by  the within-populations sample covariance matrix.  After the effects of  scale have been eliminated, linear combinations are made of the characteristics .  The coefficients are calculated in such a way as to maximize the  variation between populations, while the variation within populations is held constant.  This linear combination is the f i r s t discriminant score.  26 Then another linear combination is formed not only to have the same properties, but also to be uncorrelated with the f i r s t combination. the second discriminant score.  This is  Each subsequent linear combination is  formed in an analogous manner, in such a way as to be uncorrelated with every previous score. The result of these calculations is an n-by-n matrix of  coefficients.  A discriminant score of the form: <=^  X  +  <=<  il 1  X  +  . . . . . . .  o  . . . . . . .  12 2  o<  X  S3  Z  in n  is made, with each of the n columns of coefficients,  i  where x , x , x  1 2 x are the data of each of the a, characteristics, and ^ , n il a  „ c x L  r e the n. elements of the ith column of  ...  3 << , 12  coefficients.  in The set of n. scores so obtained Is called a set of discriminant functions. The distance function between two populations, as calculated in this program, Is the sum, from one to n, of the squares of the differences between the corresponding elements of each:  2  2  n i - .1  i  i  The square root "of this value is used in constructing models of the results of a distance function analysis. There are a number of limitations to the discriminant and distance function analysis.  The within population dispersion matrices are assumed  to be homogeneous, since the rotation of coordinate axes is not possible otherwise.  Tests of this homogeneity are not well developed.  Snyder's  program is organized to test separately for significant differences in the variances and correlation coefficients,  which is not quite the same thing  - 27 -  as testing for homogeneity of covariance„  If i t is decided that the matrices  are heterogeneous, there is no recognised course of action open which provides any estimates of the errors involved. The degree of dependence between the characteristics is defined by means of the correlation coefficient.  This coefficient is easily per-  turbed by small sample size and small size range of data values.  It is  also possible to conceive of situations where a high degree of dependence is not shown by this coefficient.  For instance, two variates might be  positively correlated for low values and negatively correlated for high values.  They would be highly dependent, but the overall correlation coef-  ficient would be close to zero. Most of the difficulties arise from the fact that the technique was actually developed for a situation where the characteristics of the populations involved were known in the beginning, where the populations were very similar, and where the samples were very large (Mahalanobis was studying the differences between human races in India).  In most of the appli-  cations of the technique, including this study, the purpose is to determine the characteristics of populations which are unknown, where the populations may not be very closely related, and where the samples tend to be small. Statistical methods Snyder's program was designed to deal with 10 characteristics, and 35 samples of up to 25 individuals each.  In every parameter these were  fewer data than those obtained by the writer, but limitations of this order were imposed by the capacity of the IBM 1620 computer a v a i l a b l e „  The pro-  gram handled raw data only, and calculated discriminant functions of samples and individuals, together with the distances between samples.  It included  - 28 -  corrections for missing data, f a c i l i t i e s for testing differences between variances and correlation coefficients,  and for determination of the rela-  tive contributions to discrimination of the various characteristics used. The method of measuring the characteristics used is given on page 5• Preliminary calculations indicated that variances were related to the square of the mean in each characteristic.  This showed that a logarithmic  transformation of the data was required, and this feature was added to the program. A preliminary check was made against the possibility of unusual relationships between characteristics which, might invalidate the s t a t i s t i c a l analysis.  Most of the data were plotted against each other on double loga-  rithmic graph paper.  Only simple regressions were found to be involved.  Next, testing for homogeneity of covariance was undertaken.  First,  the variances and correlation coefficients were tested for significant differences. were examined.  Second, the sampling distributions of these two statistics The differences found were so great, and the sampling dis-  tributions so irregular, that there seemed to be l i t t l e question that heterogeneity of covariance exists.  It was decided that a discriminant function  analysis would s t i l l be the best method of dealing with the data, but that external tests of significance, using duplicate samples, should be used instead of the internal tests commonly used (Nash and Jolicoeur, ms., p. 10). Two basic assumptions were made about the sampling.  It was assumed  that every sample consisted of only a single species, since care was taken to obtain each sample from what appeared to be a single population.  Since  no population studied showed i t s e l f in general to be shorter lived than  one year, i t was also assumed that plants occupying the same locality at different times of year belonged to the same population.  In other words,  i t was assumed that there is no seasonal succession of Alaria, species in the same locality. First run It was necessary that the 35 samples, to be chosen from those available, should be stratified in some way so that a bias should not be introduced in favour of those areas where the writer had been able to collect the most samples.  Since the Pacific samples were more diverse than the Atlantic  ones, "the question arose whether the stratification should give equal representation to suggested species, favouring the Pacific, or equal representation by area, giving a roughly equal number of samples from both oceans. It was decided to give representation by species in the f i r s t run, as this would give the best approximation of the between sample covariance matrix foi> the whole genus.  Descriptive data concerning the samples used are  listed in Appendix II, Table 1. The f i r s t run was designed to do two things.  F i r s t , a number of du-  plicate samples were included, both members of which appeared to belong to the same taxon but which were subject to different environmental conditions. 2.  The data of these duplicate samples are shown in Appendix II, Table  This duplication was intended to give both an estimate of the effects  of environment, and an indication of what distance between populations might be significant for taxonomic purposes.  Second, i t was intended to  test the reality of the writer's subjective estimates of the limits of the various species.  If the species are distinct entities without, many inter-  mediate forms, the several samples from each one should appear as a cluster in the distance function analysis.  This was especially l i k e l y , since a bias  - 30 towards clustering was probably introduced in stratifying the samples by species, and by using duplicate samples from the same species„  A lack of  clustering, with the species s t i l l occupying discrete parts of hyperspace, would indicate that the species are not distinct entities, and.that there are complete ranges of intermediate forms.  A division drawn between two such  species would be arbitrary in that i t would not represent a natural discontinuity in form between them.  An arbitrary division is meaningful i f i t  provides a workable means of separating two species which are shown to be different from one another by some other criterion.  However, If samples  thought to be derived from the same species show no tendency to occupy adjacent positions in hyperspace, no consistency is apparent between statist i c a l and other c r i t e r i a used in distinguishing the species.  A division  based on statistical c r i t e r i a would be shown to be meaningless as well as arbitrary. The results from the f i r s t run (Appendix II, Table 3) are approximated in Figure 13 by plotting the f i r s t three (largest) discriminant scores against one another.  It is apparent that clustering is the exception rather  than the rule, since It is evident only with 4. nana and A. crispa. However, the species do occupy discrete sections of space, so the divisions between them appear to be arbitrary but meaningful.  Clustering is a sub-  jective concept for which there is no statistical test (Rao, 1952, p. 351). The lack of clustering is displayed in Figure 14, where the various distances obtained in the f i r s t two runs are plotted.  If clustering were pre-  sent one would expect a bimodal distribution, with one peak for the withincluster distances and another peak for the between-cluster distances. is evident that the distribution is uniraodal. The results from the duplicate samples (Appendix II, Table 2) show  It  - 31 that sampling and measurement errors, and differences of level and exposure to surf have relatively l i t t l e effect, while differences associated with time have a relatively large effect-  In the case of samples 23 and 26, the  possibility exists that the assumption of no seasonal succession of Alaria species in one precise locality (p„ 39) did not hold, since there are no data on the longevity of Alaria angusta.  According to the internal tests  of significance (Nash, ms , p. 9) > a l l the differences between duplicate 0  samples are significant at the 99$ level except the measurement error.  It  value of less than 5 units in this run could easily  would  be attributed to environmental e f f e c t s „ Two features of the f i r s t run required further attention.  While samples  from the same area in the Pacific tend to be close together in the distance function analysis, no such effect is evident in the Atlantic.  Also, sam-  ple No. 5 (from Attu I . in the Aleutian Islands) seemed to belong to the Atlantic group.  It was decided to investigate these points in greater de-  t a i l in the second run. SCQQntl run In the second run, the representation of the Atlantic was much increased. This required eliminating some of the Pacific samples.  Care was taken to  keep Pacific samples which appeared close to the outer edges of the hyperspace outlined in the distance function analysis.  In this way i t was hoped  to maintain much the same between-sample covariance matrix as in the f i r s t run.  The sample size was increased so as to range from 20 to 25 (mean 23),  as opposed to 10 to 25 (mean 18) in the f i r s t run. This was done partly by merging samples already shown to be very similar, and partly by adding infertile plants to the sample or by measuring more individuals.  The results of the second run (Appendix II, Table 3) are approximated in Figure 1 5 . The lack of geographical differentiation in the Atlantic, as opposed to the Pacific, and the anamolous position of sample 5 (numbered 39 when increased in size) is shown more precisely in Figure l 6 .  The D values 2  between samples common to both runs were checked and found not to differ more than 10$ between runs.  The distances calculated in the f i r s t two  runs are therefore comparable, and the flfi  value of 5 units for possible  environmental effects can be used for the second run as well as the f i r s t . Third run The third run was not parallel to the f i r s t two, but was intended to increase the capabilities of the program. Effective testing and practical use of the discriminant function technique require that, unknown individuals be identified.  These unknowns should be identified according to species,  rather than to one of the original samples„  In order to do this, a-mean dis-  criminant function, weighted for the size of the various samples used, was determined for each species from the results of the second run (Nash, mg., p. 2 ) .  Then an additional section was written for Snyder's program, in  which a distance function from the mean discriminant function of each species was calculated for the unknown individual.  The individual is identified  as belonging to that one of the ten species treated from which i t had the shortest distance. The f i r s t test consisted of identifying the individuals which had been used in the second run i t s e l f . misclassified.  Nineteen percent of these individuals were  These misclassifications involve only the variability of  the samples, appreciated subjectively in the form that "Within any large sample, an occasional, apparently divergent individual can usually be found" (p. 2 3 ) .  Of course, any failure of the assumption that a sample contains  - 33 only one species would a l s o appear as an error. be detected.  Two possible errors would not  The chance existed that the ten species treated were not  adequately sampled i n the second run so that the mean discriminant functions obtained were inaccurate.  Also, the unknown might belong to a species not  among the ten treated. In order to attack the f i r s t of these latter two possible errors, a l l the small samples known to belong to these ten species were used in a second test. Table 1.  Descriptive data of these samples are provided in Appendix II, When these samples were tested using the mean discriminant func-  tions previously used in the f i r s t test, classified.  kofy  of the individuals were mis-  The increase from 19$ to k0<$> is a measure of the inaccuracy of  the f i r s t set of mean discriminant functions.  The discriminant functions of  the individuals used i n the second test were then added to the mean discriminant functions obtained in the f i r s t test to make a second set of mean discriminant functions. nant coefficients  The original covariance matrices and discrimi-  from the second run were therefore assumed to be good ap-  proximations of those in the general population, but additional data were used to make a second order approximation of the mean discriminant functions of the ten species themselves. The problem of allowing for the possibility that an unknown belongs to a species other than those studied is a very difficult  one, since the charac-  teristics and the probability of encountering an unknown species are of necessity unknown.  The sampling distribution of the values of distances  between individuals and the mean discriminant of their own species was . studied.  It was found that 99$ of the individuals were within a Jlfi  of 7 units from the mean of their own species.  value  So an unknown which shows a  of more than 7 units from a l l ten species might be suspected of belonging  - 34 -  to yet another species.  However, this criterion was found to be relatively  unsatisfactory, since Alaria esculenta is much more variable than the other than 7 units from the mean discriminant of the  species.  species would be less significant in the case of A., esculenta than in the case of the other nine species treated. A third test was carried out on the corrected mean discriminants, using data which had not been used previously, usually because identification was uncertain.  In cases where a sample was thought to be intermediate between  two species, or a mixture of them, - classification either of the two species was considered correct.  of an individual as In the case of unidenti-  fied samples, an identification as one of the species thought to be present in the area was considered correct. 7  This test was also designed to check the r e l i a b i l i t y of different kinds of data, both of those types of data which had been used in the analysis and those types which had not been used before.  The results of this test are  shown in Appendix II, Table 4. Among the types of data used in the analysis, there were 28$ misclassifications.  This was probably not much greater than proportion in the f i r s t  test (19$) considering the .generally unsatisfactory nature of the data used in the third test.  Also, an examination of the specimens from Sitka;(sample  No. 90) and Adak (sample No. 89), identified in the test as Alaria taeniata. showed that these were probably extensions of range of A. taeniata and not misxdentifications.  However, this is an §, posteriori conclusion and the  figure of 28$ was not changed to allow for i t . Among the types of data, at f i r s t only those from fertile plants were used, except in the case of A. crassifolia. where i t was thought r e a l i s i t i c to include the infertile summer form.  Where necessary to increase sample  - 35 -  size, infertile material with well developed sporophylls was added.  In ex-  ceptional cases, infertile material with poorly developed sporophylls was also used, but new measurements of sporophylls were taken from other plants of similar size.  In these cases, the sporophyll measurements were real in  their relationships with one another, but a r t i f i c i a l in relation to other measurements of the "same" plant. were of this type.  About 3$ of the data in the f i r s t two runs  Missing data, almost exclusively angles of lamina base  (measurement No. 1 0 ) , were supplied by plotting the regression of the other data for this measurement against the data for the other measurements in the same sample, and interpolating the missing value.  Occasionally, there  was a missing block of a l l of one measurement in a sample.  In this case,  the regression was supplied from a similar sample which was not used, and the values interpolated as before.  Missing values estimated by these two  methods amounted to less than 1$ of the data in the f i r s t two runs. rarely, dried specimens were soaked in water and measured.  Very  Use of this type  of data was very limited because of the destructive effect on the specimens involved and the work required to prepare them. used in this way.  No borrowed material was  Data of this type were not used at a l l in the f i r s t two  runs, and amounted to less than 1$ of the data i n the fourth run. Among the data which were not used before, were those from dried specimens and the same measurements multiplied by a correction factor obtained by measuring the same plants before and after drying.  From this test, i t ap-  pears that the use of correction factors on measurements of dried material is valueless. limited value.  The use of uncorrected measurements of dried material has The use of sporophyll measurements from another plant when  these are missing is of doubtful value.  - 36 Fourth run It was decided to reiterate the analysis by using the results from the f i r s t three runs in planning a fourth run.  In particular, the mean discrimi-  nants for each species should be calculated directly by grouping a l l samples belonging to the same species, instead of by approximation through calculation of a weighted mean discriminant function„ It was found that a l l ten eigenvalues (components of between sample variation) of the previous runs were significant (Nash and Jolicoeur, ms», p. 3 ) .  This indicated that i t might be profitable to use more characteris-  t i c s , so the program was changed to use twelve instead of ten.  These were  the f i r s t twelve measurements defined on page 5. A l l samples assigned to the same species from the results of the f i r s t three runs were grouped together so that i t was possible to obtain a large statistical sample of each species.  The program was changed to increase the maximum number of individuals  per sample to 100.  Some thought was given to the possibility of rejecting  aberrant data, but finally i t was decided to avoid bias by going back to the original criterion of using only fertile plants, except in the case of Alaria crassifolia.  After the data concerning infertile plants were omit-  ted, only Alaria esculenta had many more than 100 individuals. was divided into three groups on a geographical basis.  This species  Plants from the .British  Isles, Norway, and the rest of the world were treated separately.  1  This  geographical grouping was a check on the degree of geographical differentiation in the species. coefficients,  The presence of three columns of discriminant  and therefore three distance values, differing in an essen-  t i a l l y random manner, served ih a rough way to compensate for the greater variability of A. esculenta,,  The increased possibility that an unknown f a l l s  the shortest distance from one of the three mean discriminants associated  - 37 with A„ esculenta. would allow for the smaller significance of a distance from the mean discriminant of 4- esculenta. (p. 33). The distances between the various species calculated in the fourth run are represented in Figure 17-  These are not comparable to previous dis-  tances, since they are the sum of twelve, not ten, values. A test, using the data used in the fourth run i t s e l f , produced 13$ misclassifications.  A final test, using the 99 surplus fertile plants not  used in the fourth run, produced the misclassification of one f i f t h of the individuals used. The sample means, the within-species and the between-species covariance matrices, the variance matrix, the distance matrix, the eigenvalues, and the discriminant coefficients II, Table 5.  obtained in the fourth run are listed in Appendix  A l l these statistics are in base 10 logarithms.  Use of the discriminant and distance function It is more convenient to use these statistics for identification of an unknown with the help of the computer program written for the purpose.  How-  ever, i t is possible to make a limited number of the calculations manually, with a desk calculator.  One should f i r s t make the f i r s t 12 measurements  described on page 5 on a wet, f e r t i l e plant, and convert these measurements to base 10 logarithms.  Then take the value for measurement No..1 and  multiply i t by the f i r s t coefficient of the f i r s t column in the table of discriminant functions (Appendix II, Table 5).  Multiply measurement No. 2  by the second coefficient in the same column, and so on. values obtained for the f i r s t column together.  Add the twelve  Then take the value for  measurement No. 1 again, and multiply i t by the f i r s t coefficient in the second column, and so on.  The resultant set of 12 values, one corresponding to  each column of coefficients,  is the discriminant function of the unknown.  - 38 The sum of the squares of the differences between each element of the discriminant function of the unknown and the corresponding element of the mean discriminant function of the species i 6 the distance function or  of the  2  individual from the species.  This D should be calculated between the un-  known and each of the nine species and the three sections of Alaria esculenta treated.  The unknown is identified as belonging to the species from which  i t has the least distance.  There is no sure way of coming to the conclusion  that the unknown belongs to some other species, not treated statistically here.  In the f i r s t test of the fourth run, 99$ of the individuals used  were within a distance (ir^) of ho units from the mean discriminant function of the species to which they belonged.  So i f an unknown is more than kO  units from a l l ten species, this is a clear warining that another species may be involved.  A conclusion to this effect could not, of course, be based  on a single individual.  Before a new species could be included in a dis-  criminant function analysis, a large s t a t i s t i c a l sample of i t should be collected, and the discriminant and distance functions recalculated, including the new sample. Decisions based on the s t a t i s t i c a l analysis The results of the analysis were not in themselves conclusions, but formed the basis on which decisions were made.  The selection of samples and  characteristics at the beginning probably caused the final results to reflect the writer's i n i t i a l evaluation of species boundaries, based on widespread field observations and study of herbarium specimens.  In this study, statis-  tics were used to reduce a large amount of data to manageable size, and  to  ensure that the decisions made should be consistent with one another and capable of being communicated in an unambiguous form.  The inclusion of sample No. 39 (Appendix II, Table l ) from the Pacific Ocean in the population of Alaria esculenta from the Atlantic Ocean was the basis for the decision to include a l l specimens from the intervening Arctic in Alaria esculenta. except for a few collections which were obviously distinct .  These latter collections formed Alaria p y l a i i and Alaria grandifolia.  which also showed inrboth:Atlantic and Pacific. The lack of geographical differentiation in the Atlantic, taken in conjunction with the results of Sundene's work (I962) indicated that there is only one species throughout most of the temperate Atlantic.  Alaria p y l a i i  probably does not grow in Scandinavia and Alaria p y l a i i sensu Kjellman (189O, p. 20) is probably at most an ecotype and has no taxonomic- significance. Examination of the plentiful herbarium material from the Kurile Islands available at Hokkaido University indicated that there is tween 4. crispa in the north and A. angusta in the south.  a gradation beHowever, the dis-  tance function analysis showed that A. crispa was much closer to A. nana than to A. angusta.  Since field observations indicated that A. crispa and  A. nana are distinct, i t was decided to keep 4. crispa and 4. angusta distinct also, at least u n t i l f i e l d observations can be made in the Kurile Islands. It had already been noted in the literature (Yendo, 1919, p. 96) that 4. praelonga and 4. marginata are very closely related species.  A careful  examination of the distance functions between the various samples indicated that they can be separated, though the separation is made at an arbitrary point in what appears to be a complete gradation of form.  Since the two  species'have quite distinct geographical distributions, i t was decided to keep them separate.  The possible biological significance of differences of variance between species, and of the lack of geographical differentiation in the Atlan* t i c , is discussed in the conclusions.  41 Limits of the genus Alaria Greville Alg. B r i t . 25. I.83O. "Frond membranaceous, furnished with a percurrent cartilaginous midrib, the stem pinnated with distinct leaflets." The name Alaria is conserved against Musaefolia Stackh., 1809, p. 53 (Lanjouw, 1961, p. 210).  Fugus esculentus L», 1767, p. 135; is cited as  the type species of both genera by Silva (.1952, p. 259). The tax.onomic limits of the genus AlarjLa have remained essentially unchanged since Greville, and the description need not be amplified to exclude new genera discovered since his, day. One doubtful case is Pleuropterum Miyabe et Hagai, 1932, p„ 127.  It  is similar to Alaria. except that some of the lateral branches are sterile, and some have percurrent midribs.  Generic distinctions are largely subjec-  tive, so there are no clear c r i t e r i a for including or excluding this genus from Alaria.  Only a few specimens have been collected, so the material on  which a decision must be made is limited. It is unwarranted to base a generic distinction essentially on a single character.  Even in the few specimens collected, this character is shown to  be variable; in some specimens the number of later branches different from Alaria sporophylls is much smaller than in other specimens.  If this one  character is ignore*d, Pleuropterum paradiseurn appears to be identical with Alaria dolichorachis f. longipes Miyabe and Pleuropterum fasciculatum Yamada identical with some specimens of Alaria angusta.  Thus i t is possible  that this character may be a relatively simple mutation which can arise in several species of Al&ria. so that the several entities of Pleuropterum are more closely related to certain species of Alaria than to each other.  -  k2  -  It was decided to reduce Pleuropterum to synonymy under Alaria.  When Pleuro-  terum is synonymized with Alaria. Greville's generic description can he maintained unchanged.  - 4 3  Key to the species of the genus, Alaria Dimensions are taken from the wet, fertile plants of the species treated s t a t i s t i c a l l y .  Dimensions of dried specimens would be about one-  half or two-thirds of these values.  Measurements are to be made as des-  cribed on p o 5.  1.  Midrib fistulose (furnished with, air chambers) at intervals . . . . . . „ „ „ . . . . . . . . • Midrib solid throughout . . . . . . . . . . . . . . . . . . . ,, . . . 2 1 •  1. 2.  Q O  At least some lateral branches (fertile or sterile) with midrib Lateral branches entirely devoid of midr i b , usually fertile (sporophylls) . . . . . .  1 o O O  2.  A o  A. paradisea.  3. 4. 4.  Stipe more than 15.4 cm long . . . . . . 5 Stipe less than 15.4 cm long . . . . . . . . . . . . . . .. . . . 6  Sporophylls less than 2.8 cm wide 5. 5. . Sporophylls more than 2.8 cm wide  V  ..........  t o 9 «  Stipe less than 3.0 cm long Stipe more than 3.0 cm long  7. 7.  Sporophylls more than 2.8 cm wide Sporophylls l e s 6 than 2.8 cm wide  8. 8.  Stipe cylindrical, 0.3 cm or less wide Stipe sometimes flattened, 0.3 cm or  ................ ................  > « • o  4  ochotensls,,  p. 80..  A. grandifolia, p. 57. 4. taa&ifQiia, P. 71. 4. taeniata, p . 70. 7  ....  A .  marginata.  p..59.  9  . . . . . . . . . .. . . . 4. praelonga. p. 66. 4. p y l a i i , p. 68. ..........  9. 9.  Sporophylls less than 4.5 cm wide Sporophylls more than 4.5 cm wide  10.  Sporophylls tending to be fasciculate with more than one rank on each side Sporophylls one-ranked at point o f £L"fc"tjQ,Cl}ffiSH"fc>  A .  . . . . . . . . . .. . . . 8 10 ..........  1 . * «  10.  p. 65.  •  ' O Q •  6. 6.  55.  3  Stipe virtually non-existent; lower trunk bearing two ranks of persistent Gporophyll petioles or adapted haptera . . . . . . . Stipe present . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  l * •  P .  •  ' o • a  3.  f istulosa.,  > e o «  11  o o t > o © * , o e u o . « u n o o c t > o © o o o « « t ) O o o <) t> o «  12  11.  Sporophylls cylindrical in cross-section closs to the petiole, very thick and  11.  Sporophylls flattened in cross-section „ . . . ,  > * c o > o v •  A. crispa,, p„ 49. A. angusta. p . 45.  p. 6 2 .  12 o 12.  Rachis evenly tapering distally . . . . . . . . . . . . . . . A. nana» Rachis not tapering although sometimes constricted at transition zone . . . . . . . . . . . . . 13  13.  Two kinds of sporophylls; one, most common in summer, dark, thick, sterile; the other, more common in winter, lighter in colour, thin, fertile . . . . . . . . . . . . . . . . . . . A. crassifolia. p„ 4 7 . Sporophylls usually a l l of one kind; i f some are thickened, of unequal thickness in different parts . . . . . . . . . . . . . . . . . . . . . &. esculenta. p. 51.  13.  - 45 -  Alaria angusta Synonyms:  Kjellman, Svensk Vetens.~Akad„ Hand. Alaria laticosta  Illustration:  23.J38, 1889.  Kjellm., 1889, p.. 40.  Figure 32.  Notes on the type: The type locality was stated to be Bering Island (Commander Is., U.S.S.R.).  Seven sheets from Kjellman's collections, one at Lund (LD 2080)  and six at Uppsala (UPSV), are labelled as Alaria angusta from Bering Island, a l l dated 1.4-19 August, 1879, in his hand. also annotated "orig,. exemplar". latter two.  Two of the Uppsala sheets are  The lectotype here designated is one of the  The specimen may be identified by referring to Figure 32.  Geographical distribution In general, Alaria angusta is found in Japanese waters, usually on exposed beaches north and east of Cape Erimo, Hokkaido.  The type locality  would appear to be near the northern limit of the species.  Kawabata's record  from Shikotan I. (193&, p. 204) is almost certainly correct in part.  E . S.  Zinova's reports from Cape Ustrecbni, Tyulen'oi Rook, Ukhidnoi Reef, Medni I . , Preobrajeniya Bay, and Tepsikova Reef, in the Soviet Pacific (1929, p. 30; 1940, p. 193; 1954b, p. 330), could not be verified. both A. angusta and A. erispa.  They probably refer to  If they could be authenticated,  these reports  would make the general distribution of the species more consistent with the type locality.  The distribution is illustrated in Figure 18.  Relation to other species and the environment Without observations on the main distribution of the species, in the South Kuriles, i t is difficult to form a clear picture of the  relationships  between this species and other species, and between this species and its environment.  In Japan, A. angusta is characteristic of beaches exposed to surf,  k6 with an upper limit f a i r l y high in the low intertidal zone.  I t seems to be  an ill-defined species, merging into A„ c r a s s i f o l i a near Cape Erimc, into A. ochotens_is in Saghalien, into A. paradigsea in the Middle Kurile Islands, and into A, crlspa in the North Kurile Islands. Representative specimens examined NORTH KURILE ISIANDS: Odomari, Onnekotan I . , Aug. 15, 1935? 4= taeniata. Nagai (Fac. Ag., Sap.). MIDDLE KURILE ISLANDS: Urup I., 1892, Yendo 225B (Fac. Ag., Sap.); Tokotan, Urup I . , Aug. 6, 1930; Nagai (Fac. Ag., Sap.); Yema, Urup I . , July, 1933, Yamada (SAP 026607), SOUTH KURILE ISLANDS: Shana, Etorup I., Aug., I89O, S. Fu.limura (Fac Ag., Sap.); Chikohai, Etoruf I . , Aug. 8, 1931, Nagai (Fac. Ag., Sap.); Cape Butsushiyo, Etorofu I . , Sept. 30, 1934, S. Nagamine (Fac. Ag., Sap.); Kamuiwakka,- Etorofu I . , July 27, 1894, Yendo 226A (Fac. A g . . Sap.); Chinomichi, Kunashiri I . , Aug. 28, 1925, A. praelonga. J . Tokida (Fac. Ag., Sap.). SAGHALIEN: Moiretomari, W. Coast, June 13, 1907, A. dolichorachis. I. Miyabe (Fac. Ag., Sap.); Shiranushi, Aug I 7, I906, A. dolichorachis. Miyabe (Fac. Ag., Sap.).  /  - 47 Kjellman, jLn Kjellman et Petersen, Jap. Lam. 276. I.887.  Alaria crass i f olia Synonyms:  none.  Illustration:  Figure 33•  Notes on the type: The type locality was stated to be Hakodate, Hokkaido, Japan.  There  is one sheet in Kjellman s collections (OPS?), labelled "Alaria crassifolia ,  Kjellm. Insula Jezo 18 ^ 8 l " in his hand.  The one complete specimen on this  sheet is here designated as the lectotype, and is illustrated in Figure 47. Geographical distribution: Alaria crassifolia has a d e a r l y defined and restricted distributionon the Pacific coast of the Home Islands of Japan, from Gape Erimo south and west to Miyagi Prefecture in Honshu, A few individuals in the South Kurile Islands also appear to belong to this species.  The distribution is i l l u s t r a -  ted in Figure 19. Relation to other species and environment: In its Home Island distribution, A. crassifolia can survive sea temperatures (Figure 9) well above the l6°C limit suggested by Sundene (.1962). During the summer, i t forms very thick, dark coloured, infertile sporophylls, which possibly store food reserves.  The summer form of A. c r a B s i f p l i a is very  similar in morphology to A. esculenta,. but in the latter species the development of thickened, sterile sporophylls in summer, and also resistance to high sea temperatures, is not so great.  During the winter, 4 . crassifolia forms  wide, thin, fertile sporophylls which, i f the other kind of sporophylls is not present, make the plant appear very similar indeed to A. praelonga.  Such  a sporophyll is shown on the left hand side of Figure 33. Since the most distinctive characteristics of A. crassifolia are so closely connected to an unusual range of sea temperature, the identification of certain individuals  - 48 from the South Kurile Islands, where this environment i s absent, must remain somewhat doubtful.  One of these plants i s shown on the right hand side  of Figure 33. These individuals may actually by A. esculenta.  In any case,  there i s l i t t l e doubt that A. esculenta and A. crapsifqlia are very closely related, and that their distributions are nearly continuous with one another in the northwest Pacific. Representative specimens examined; SOUTH KURILE ISLANDS: Chiboi, Shikotan I., cast, ashore. July 30, 1934, Alaria angusta. Nagal (Fac. Ag., Sap.); Arimoe,. Etorof I„, Aug. 8, 1931, A. macroptera. Nagai (Fac. Ag., Sap.); Kamuiwakka, Etorofu I., July 27, 1894, No. 224, A. macroptera. (Fac. Ag., Sap.); Rubetsu, Etorof I., Aug. 7, 1930, Nagai (Fac. Ag., Sap.). HOKKAIDO: Muroran, A p r i l 15, 1953, E. Y. Dawson 11729 (UC H 7 2 9 ) . HONSHU: Shimofuro, June 10, 1915, K. Tago (Tl).  - 49 Alaria crispa  Kjellman Svensk. Vetens.-Akad. Hand.  23'37. 1889-  Alarja lancegjLata K j e l l m 1 8 8 9 , p. 39.  Synonyms:  Notes on. -yig, type: The localities were stated to be St. Lawrence Bay, St. Lawrence I. (Bering Sea).  These were probably two l o c a l i t i e s , St. Lawrence Bay being  on the Siberian mainland.  There are seventeen sheets in Kjellman's collections  identified in his hand as A. crispa.  Three sheets.(one at Lund (LD 208l) and  two at Uppsala (UPSV)) are from St. Lawrence Island;  fourteen sheets ( nine  at Stockholm (S) and five at Uppsala (UPSV)) are from St. Lawrence Bay. One of the specimens on the sheet LD 2081, collected I/8/.I879, is here designated the lectotype.  It is illustrated in Figure 48.  Geographical distribution: Alaria crispa occurs frequently in the North Kurile, Commander, and Aleutian Islands.  Its distribution is illustrated in Figure 2 0 .  Relation to other species and environment: Alaria crispa is a name which has been virtually unused since its description by Kjellman, but the original collections are similar to a well defined series of populations observed by the writer in the Aleutian Islands. 4. crispa occupies a distinct zone in the mid-intertldal zone, closely associated with Laminar.ia longipes Bory.  A. crispa is most commonly found on beaches  exposed to surf, but may also occur in sheltered bays, in which case the sporophylls tend to be v/ider and thinner as illustrated on the right hand side of Figure 34.  It is probable that the A. crispa. zone is much lower in the  intertidal zone in northern l o c a l i t i e s , where grinding by ice is a feature of the environment. When 4 . crispa is established as a separate entity, i t becomes apparent  - 50 that a portion of the collections identified as A. angusta by Japanese workers in the Kurile Islands  actually belongs to 4- erispa.  There appears  to be a gradation between 4 . erispa in the North Kuriles, and A„ angusta in the South Kuriles. shown in Figure 34.  An intermediate form between 4 . erispa and A, angusta is However, the populations observed by the writer at the  two ends of this series are quite distinct, as shown by the distance function analysis (Figure 13).  Since direct observations of the populations in  the Kurile Islands were lacking, i t was decided to keep the two species as separate entities and to typify them so as to preserve current usage in Japan. Representative specimens examined; ALEUTIAN ISLANDS: Agattu I . , lB&k, 4 . dolichorachis. Townsend 5J5^ (UC 96558); Atka I . , May, 1931, A. dolichorachis. I . Kobayashi (SAP); Amaknak I . , June 23, 1899, A. dolichorachis. SetcheJX 32ia (UC 96608). COMMANDER ISLANDS: Bering I . , 14-19/8/1879, 4 . lanceolatao Kjellman (LD 2083). BERING SEA: St. Lawrence Bay, 20/7/1879, Klellman (S). NORTH KURILE ISLANDS: Tenjin-iwa, Shumushu I . , July 19, 1932, 4. angusta. Nagai (SAP); Kataoka-wan, Shumushu I . , July 28, 1930, 4 . angusta. Nagai (Fac. A g . , Sap.); Kakamubetsu, Paramushir I . , July 24-26, 1930, 4 . angusta, Nagai (Fac. A g . , Sap.); Cape Sekine, Araito I . , Aug., 1926, 4 . taeniata. &. Itoh and G,. Kumori 1510 (Fac. A g . , Sap.). MIDDLE KURILE ISLANDS: Yamato-wan, Matsuwa I . , Aug. 14, 1935, 4 . angusta. Nagai (Fac. A g . , Sap.); Minamiura, Ketoi I . , Aug. 20, 1929, A. angusta, M. Tatewaki, K>. Takehashl (Fac. A g . , Sap.); Broughton-wan, Shimushir I . , July 22, 1930, 4 . angusta. Nagai (Fac. A g . , Sap.); Mishima, Urup I . , Aug. 8, 1935, 4 . angusta. Nagai (Fac. A g . , Sap.). SOUTH KURILE ISLANDS: Toshirari, Etorof I . , Aug. 7, 1931, 4 . angusta. Nagai (Fac. A g . , Sap.).  -  Alaria esculenta Synonyms,:  (L„) Greville  51  Alg. B r i t . 25.  I83O.  Alarum, delisei. Bory, 1826, p. 194. Alaria delisei (Bory) Grev„, 1830, p. xxxix. Q£gya. delisei (Bory) Trev., 1845, p. 28. Alaria doliphorachis. Kjellman, 1883a, p. 217. Fucus esculentus L . , 1767.9 p. 135. Geramium esculenturn (L.) Stackh., 1797, p. 24. Musaefolia esculenta (L.) Stackh., 1809, p. 53. Qrgyia esculenta (L.) Stackh., 1816, p. 8. Laminaria esculenta (L.) G. A g . , 1 8 1 7 , p. 13. Phasgonon esculentum (L„) Gray, 1821, p. 385. Agarum esculentum ( L . j Bory, 1826, p„ 194. Podopteris esculenta ( L „ ) De l a P y l , , 1829, p. 23. Qrava esculenta (L„) Trev., 1845, p. 28. Alaria esculenta f. fasciculata Stroemfelt, 1886, p. 38. Fucus escalentus v a r m i n o r Turner, 1802, p. 1G4. Phasganon esculentum^ minus Gray, 1821, p. 385. Phasganon macroxiterum Ruprecht, I85I, p. 353. Alaria, msxpptera (RuprJ Yendo, 1919, p. 79» A l a r i a musaefolia (De l a P y l . ) J . A g . , 1872, p. 23.  LaMaaxia. musaefolia. De l a P y l . , 1829, p. 31.  A l a r i a eBcu3,gjRlift f . mqsaefolja (De 1883a, p. 212. Laminaria musaefolia/s remotifolia Alaria oblonga Kjellman, 1883a, p. Fucus teres Good. eiWoodw., I797, Illustration:  l a P y l . ) Kjellman,  De l a P y l . , 1829, p. 35. 220. p. 140.  Figure 35.  Motes on the type: The type l o c a l i t y was stated to be the Atlantic Ocean, and probably is somewhere in Norway. The only specimen of Alaria i n the Linnaean Herbarium cannot be proved to date from the time of the description, but i t does have the annotation "esculentus" in Linnaeus' hand, and agrees in a l l p a r t i culars with the description.  Since there does not appear to be any other  suitable specimen, this one (LINN 1274.63 (Savage)) is here designated the lectotype.  If i t should be proved that this specimen does not date from I767, i t  would be a neotype instead of a lectotype.  It is illustrated in Figure 49.  Geographical distribution: The principal area of distribution of A„ esculenta. where i t is the  - 52 only common species of the genus, is in the temperate waters of the Atlantic, that i s , in Norway, the British Isles, Iceland, the Canadian Maritime Provinces, and New England.  The species also occurs occasionally in the Arctic,  the Aleutian Islands, and the Kurile Islands, as Is discussed below.  The dis-  tribution of A. esculenta is shown in Figure 21. Relation to other species and environment; In.the Atlantic, A. esculenta occurs south of the limit of winter sea ice, and north of the 16 C summer isotherm of sea temperature, in areas of extreme to moderate exposure to surf.  In areas of extreme exposure to  surf, the Alaria zone extends from, the sublittoral into the mid-intertidal zone (Powell and Chamberlain, 1956, p. 2).  In areas of moderate to slight ex-  posure to surf, the upper limit of ALaria is close to low water mark. Alaria esculenta probably presents•more (problems in;classification than any other species of Alaria.  The problems considered here are the re-  lationship between the North American and European populations of the species, the identity of Alarja p_yja.il as this name is used by Scandinavian authors, and the relationship between A. esculenta In the Atlantic to certain Arctic, North Pacific, and Japanese populations. No meaningful way was found to separate the populations of the various areas of the Atlantic on a morphological basis (see page 32).  These  areas are quite widely separated geographically, but l i t t l e divergence i £ apparent in the respective populations of A. esculenta. The writer did not find any populations in Europe which could be . attributed to A., p y l a i i sensu Kjellman, but could only find a tendency to the formation of a sheltered water form with a longer stipe and a more cordate lamina base than exists in the exposed coast form.  Sundene's transplant  experiments (1962, p. 160) showed that plants transplanted from a moderately  -  53  -  exposed coast to a sheltered one quickly developed wider blades.  The writer's  experiments in Scot-land (p. 7) show that there is a continuous growth of the blade at the lateral margin.  If the sheltered water habitat causes either  an increased rate of growth or a decrease^ rate of erosion, a wider blade would result.  Thus, It appears that a wide blade is a reflection of a r e l a -  tively simple environmental effect on the plant and is of no taxonomic value. There is l i t t l e evidence that more than one species of AJ.&riS. occurs on European Atlantic coasts. The distance function analysis indicated that one of the samples from the Aleutian Islands should be included in the Atlantic group of Alaria esculenta (p. 3 9 ) .  A further examination of this sample, and Of a number of  individuals among herbarium specimens, showed that indeed these could not be distinguished from A. esculenta. Only a very few specimens, and no field observations at a l l were available from the Arctic.  It appeared reasonable, considering that the Arc-  t i c l i e s between the populations of A. esculenta in the North American Atlantic,  the European Atlantic, and the Aleutian Islands, to assume that A. escu-  lenta grows in the Arctic also.  So, a l l Alaria plants from the Arctic which  are not obviously different from the populations in the Atlantic were i n cluded in A. esculenta.  A. esculenta Is the type species, so i t is also the  correct policy, from the point of view of nomenclature, to Include in i t a l l populations which in the present state of knowledge cannot be defined as separate species. The samples of Alaria crassigpljLa also were found in the distance function analysis to be very closely related to A . esculenta.  A further s i -  milarity between the two species is the tendency of A. esculenta in some Norwegian fjords to form thick sporophylls in summer and thin ones in winter  (Sundene, 1962, p. 160).  4° crassifolia i s , however, kept separate from A.  esculenta. as explained above (p.  48).  Representative specimens examined: FRANCE: $Le de Batz, Roscoff, 16 Aug.. 1898, A. Vickers and I . Karsakoff (PC); Cherbourg, A p r i l , 1918, S_. Corbilre (PC); Le Conquest, 10 Oct., 1850, Gomont (PC); Ouessant, 10 Oct., 1925 (PC)._ BRITISH ISLES: Anglesea, Hugh Davies (K); Kings Cove, l 6 June, 1798 (K); St. Michael's Mount, June, 1799 (K); Isle of May, Greville (BM); Dunbar, GrsyjLIle (ED); Aberdeen, 1884, Dickie (BM); Berwick, 1884..Batters (BM); Kynance Cove, Ed.. Forster (BM)$ St. Ives, June, I883, F_. W. Smith .(BM); S k a i l l , Orkney, 1838, J . H. Pollexfen (BM); Rockall, 28 June, .1921 (PC). NORWAY: Mandal, 30/7 ./l86l."wittrock (UPSV); Aalesund, 1872, A. pylaii,, Kjellman (LD 2099); Lodingen, Nordland, 4/1882, A. membranacea. Foslie (UPSV). FAERGE IS.: L i l l e Dimon, 4/8/1902, Boergesen (BM); Skulefjord, 28/IO/I897, H. Jonsson 46 (C). ICELAND: Karlskali, ReyWf jorSur, 29 June, 1883, Stroemfelt (S). GREENLAND: Sukkertoppen, I87O, A. p y l a i j , Berggren (LD 2098); Godhavn, 26 July, 1886, A. p y l a i i , Rosenvinge 366 (C); Cape Hope, 30/6/33, A. pylaji f. membranacea (S. Lund); Kangerdlugsuak, A. p y l a i i (C). CANADIAN ARCTIC: Whale Sound, Davies Straits, cast up_, Aug. 5, 1852, Dickie (BM); Queens Channel, Northumberland Sound, Lat. 76 52' N . , 10 Jaly, 1853, A. pylajLj, D. L.(K). SOVIET ARCTIC: Pitlekaj, l/j/lBj9, A. dolichorachis. Klellman (LD 2251); Koljutschinfjorden, 21/6/79, A. elliptica., Kjellman (S): Irkajpl, Sept. I3-I8/8/78, A. oblonaa. Klellman (LD 2084). KAMCHATKA, OKHOTSK SEA, and SAGHALIEN: Pankara, A. dolichorachis. Miyabe (Okamura); Drankinski, 12/ 7/27, A. dolichorachis. Miyabe (Okamura); Ozernaya, Kamchatka, July 2 9 , 1930, A. tenuifolia. £ . Umeno (Fac. Ag., Sap.); Gape Notoro, West Coast, Saghalien,,July 19, 1906, A. ochotensis. Miyabe (Fac. Ag., Sap.). NORTH KURILE ISLANDS: Chitose-wan, Paramushir, July 2 9 , 1930, 4 . dolichorachis. Nagai (Fac. Ag., Sap.); Kataoka-wan, Shumushu I . , July 2 8 , 1930, 4. praelonga. Nagai (Fac. A g . , Sap.). MIDDLE KURILE ISLANDS: Broughton Bay, Shimushlr I . , 22/7/1930, 4 . tenuifolia f. iyjeicj., Nagai (Fac. Ag., Sap.). SOUTH KURILE ISLANDS: Rubetsu, Etorofu I . , 1902, 4 . dolichorachis, K. Isa&Q. (Fac. Ag., Sap.); Toshirari, Etorofu 1., Aug. 7, 1931, A. praelonga, Nagai (Fac. Ag., Sap.).  -  Alaria fistulosa  Postals et Ruprecht  55  111.  -  Alg.  11.  l84o.  Orgya f i s t u l o s a (P. e t R.) Trev., 1 8 4 5 , p. 2 8 . Phasganon fIstulosum (P. et R.) Rupr., 1 8 5 1 , p. 3 5 5 . Alaria fistulosa f . platyphylla Setchell, i n Collins, Holden, et. S e t c h e l l , 1 9 0 1 , XLII.  Synonyms;  Illustration:  Figure 3 6 .  Notes on the type: The l o c a l i t i e s were given as I l l u l u k Bay, Unalaska I., r e f e r r e d to Merterns; Kodiak Island, referred to Kastalsky; and Ka.raginsk I., r e f e r red to Postels.  The Mertens' reference  servations only.  (I.829)  i s probably based on h i s ob-  A number of specimens consisting only of portions of lamina  were d i s t r i b u t e d by Ruprecht from the S t . Petersburgh Herbarium, now  at Kew  ©ne each are  (K), Farlow (FH), Dublin (TGD), .Land (ID), and P a r i s (PC), and  at Stockholm (S).  two  These are u s u a l l y i d e n t i f i e d as "Phasganon fistulosum  (PR)" i n Ruprecht's hand.  One of the Stockholm specimens has Ruprecht's.an-  notation " A l a r i a f i s t u l o s a Post. Rupr. Unalascha Herb. Petrop. Specia p l e t a i n herbario nondum existunt".  com-  This annotation indicates that there i s  no use i n looking f o r a better specimen at Iendlngrad (LS), so the Stockholm specimen described i s designated here as the leetotype. an incomplete  The designation of  specimen i s j u s t i f i e d here, since the character of the species  i s f u l l y evident i n the lamina.  The leetotype i s i l l u s t r a t e d i n Figure 5 0 .  Geographical d i s t r i b u t i o n : A l a r i a f i s t u l o s a occurs commonly i n the K u r i l e , Commander, and Aleut i a n Islands, and i n the Gulf of Alaska. Figure  The d i s t r i b u t i o n i s i l l u s t r a t e d i n  22. Relation to other species and the environment; A l a r i a f i s t u l o s a i s one of the most e a s i l y recognised species of  A l a r i a . because of Its u n f a i l i n g l y f i s t u l o s e midrib, i t s gregarious nature,  - 56 and its unique ecology.  If the fistulose midrib is ignored, the distance  function analysis shows that i t is very similar to A. praelonga, with which i t is commonly associated. Alaria fistulosa is found growing In the sublittoral zone, in beds fringing coasts exposed or slightly exposed to surf.  The chambers of the  midrib enable the lamina to float along the surface of the water, in a habit similar to Nereocystls, Maerocystis. and Pelagophycus. This floating habit may have three effects on the biology of the species, because of the fact that floating plants can more easily be transported over greater distances than is the ease with less buoyant individuals. The species should be able relatively easily to repopulate areas where i t has been wiped out by temporarily unfavourable conditions, drift specimens may be found far from the nearest attached plants, and gene flow should be maintained between widely separated populations, Representative specimens examined; SOUTH-EAST ALASKA: Glacier Bay, Saunders 80 (FH). GULF OF ALASKA: Haryerster I . , Uyak Bay, Aug. 20-23, l999> A. fistulosa f. etenophylla. Setchell and Lawson 5117 (UG 966OO). ALEUTIAN LSLANDS: Unalaska, June-Aug., 1899, Setchell and Lawson 1273 (UC 966OI); Kyska I., I894, (UG 96604). COMMANDER ISLANDS: Bering I . , l4~19/8/l879, K-iellman (UPSV). NORTH KURILE ISLANDS: Kakamubetsu, Paramushir I . , July, 1930, Nagai (Fac. Ag., Sap.); Murakami-wan, Paramushir I . , July 27, 1930, Nagai (Fac. Ag., Sap.); Kosikagawa, Shumushu I . , July 19, 1932, Nagai (Fac. A g . , Sap.); Minami-ura, Alaid I . , Aug 1, 1930, Nagai (Fac. A g . , Sap.).. MIDDLE KURILE ISLANDS: Broughton Bay, Shimushir I . , 22/VIl/l930, Na^ai, (Fac. Ag., Sap.); Mishima, Urup I . , Aug. 8, 1935, Nagai (Fac. Ag., Sap.); Tokotan, Urup I . , Aug. 6, 1930, Nagai (Fac. A g . , Sap.); Yamato-wan, Matsuwa I., Aug. 14, 1935, Nagai (Fac. Agi, Sap.); Usshir. I . , Aug. 12, 1935, Nagai (Fac. A g . , Sap.). SOUTH KURILE ISLftNDS: Chikohai (E. C . ) , Etorofu I . , Aug. 8, 193^ Nagai (Fac. Ag., Sap.); Moyoro, Etorofu I . , Aug. 13, 1931, Nagai 15j4 (Fac. A g . , Sap.); Shikotan sima, July, 193^, Kawabata (SAP 15616).  Alaria grandifolia Synonyms:  J . Agardh Svensk. Yetens.-Atad. Hand. 1Q_S26. 1872.  A l a r i a doljchoraebls f „ longipes Miyabe, 1928, p. 957. M a r i a p y l a i i v a r . g r a n d i f o l i a ( j . Ag.) Jonsson, 1904, p., 20.  I l l u s t r a t i o n ; Figure 34. Hotes on the type: The type l o c a l i t y was stated to be Spitsbergen, and the type c o l l e c t i o n made by Dr. Berggren.  There are f i v e sheets i n Agardh's c o l l e c t i o n s  (LD) l a b e l l e d " M a r i a g r a n d i f o l i a J „ A g . Spitsbergen, Berggren".  Since one  of these specimens (LD 2256) was c i t e d and i l l u s t r a t e d by Yendo (1919, P I . XIV),  i t appears to be the most s u i t a b l e .  I t i s here designated as the l e e t o -  type and i s i l l u s t r a t e d i n Figure 51. Geographical distribution.: M a r i a g r a n d i f o l i a occurs most commonly i n the A t l a n t i c A r c t i c ; i n Labrador, Greenland, Spitsbergen, and perhaps Novaya Zemlya.  Some i s o l a t e d  populations occur i n the P a c i f i c , i n Saghalien and the Kurile Islands. The d i s t r i b u t i o n i s i l l u s t r a t e d In Figure 23. Relation t o other, species and the environment: I t i s possible that A. g r a n d i f o l i a i s merely a large form of A . esculenta which grows i n r e l a t i v e l y cold and deep waters of the A r c t i c .  How-  ever, the form of the sporophylls appears to be d i f f e r e n t from A . esculenta. as they are often much longer, narrower, and more attenuate i n A„ g r a n d i f o l i a . Measurements of dried plants of A. g r a n d i f o l i a i n a distance function analysis indicate that i t i s d i s t i n c t from A. esculenta.  The occurrence of small  plants i n the A r c t i c which are f e r t i l e and which are i d e n t i f i e d in^this study as A . esculenta. also indicates that the two species are d i s t i n c t .  In the  present state of knowledge, the best p o l i c y i s to keep the two species separate .  - 58 The P a c i f i c Ocean d i s t r i b u t i o n , which obtains from the reduction of A l a r i a dollchorachis f . longipes Miyabe to synonymy under A„ grandifolia. i s based on a very few specimens from an area not v i s i t e d by the w r i t e r .  How-  ever, the morphological s i m i l a r i t y i s very d e f i n i t e , and a two ocean d i s t r i bution f o r A. g r a n d i f o l i a has precedent i n s i m i l a r d i s t r i b u t i o n s of A. esculenta and A. p y l a i i . Representative specimens examined; SPITSBERGEN: Treurenberg Bay, II/7/1873, K^ellman (UPSV)j Mossel Bay, Oct.Nov., I872, Kiellman (UPSV)j Fairhaven, J u l y , 1872. KJellman (UPSV); Isfjorden (Svendsen). GREENLAND: Cape Hope, 29/6/33, A. -pylaii f . g r a n d i f o l i a (S. Lund). LABRADOR: Hebron, 1955 (Wllce); S t r . of Belle I s l e , 1957 (Wilee). SAGHALIEN: Kita-Shiranushi, Prov. Honto, Saghallen, Sept., 1927, 4 . dollchorachis f . longipes. Toklda 348 (UC 543932); Hishitoma, Prov. Honto, Saghallen, Aug. 9 , 1926, Tokida (Fac. A g . , Sap.).  - 59 Alaria marginata  Postels et Ruprecht 111 . Al,g. 1 1 . 1840.  Synonyms; Alaria cjartipes Saunders, 1901a, p. 5^1. Orgya marginata (P. et R „) Trev.. 1845, P» 28. Phasganon marginatum (P. et R.) Rupr., I85.I, p. 355. Alaria valida Kjellm. et, Setchell, i n Setchell et Gardner, 1903, p. 278. Alaria valida f . lpngipes Setchell, j_n Setchell et Gardner, 1903, p. 279. Illustration:  Figure 38.  Notes on the type: The locality was stated to be i n the northern Pacific "v. g. ad Unalascha", but Ruprecht (185I, p. 355) only mentions having plants from Fort  Ross in California.  The absence of mention of material from Unalaska  in Ruprecht's detailed l i s t , and of an illustration i n Postels and Ruprecht (1840), raises some doubt that specimens were ever collected by Postels and Ruprecht.  There i s one specimen in Areschoug's collection (S), which has  a printed label reading "Herb. Acad. Petrop. California boreal Ross leg. Wosnessensky", and i s annotated in Ruprecht s hand "Alaria marginata PR (sp. 1  meliore exhiberi non potest)".. This annotation indicates that no better authentic specimen exists at Leningrad (LE). It appears impossible to prove that the Stockholm specimen i s a possible lectotype, but i t appears to be the most suitable extant for designation.  It i s clearly connected with Ru-  precht, i f not with Postels and Ruprecht.  It i s designated here as a neo-  type, and i s illustrated i n Figure 52. Geographical distribution: ^  Alaria marginata i s found i n a zone close to low water mark on v  shores with great to moderate exposure to surf, between Yakutat, Alaska, and Point Sur, California.  This distribution i s illustrated in. Figure 24.  Relation to other species and the envii'onment: Alaria marginata has a complex relationship with the two other  - 60 species commonly found in the same general area.  One of these, A. nana.  grows in a mid-intertidal zone at isolated localities where exposure to surf is severe.  Occasionally the zones of A» nana and A. marginata reach one an-  other in the vertical plane.  At Yakutat and Sitka, in southeast Alaska, the  two species remain distinct at their common boundary. At the Loran Station, Spring Island, Kyuquot, on the west coast of Vancouver I „ , B. C , they intermix.  In general, A . nana becomes more and more d i f f i c u l t to distinguish from  A. marginata towards the souther nly part.-of their (Common -range. '.. Thus- Doty's doubt (1947, p. 43) about the validity of A. nana was justified from an observation of the Oregon populations.  It only becomes apparent that the two species are  distinct in the northern part of their range, from Port Renfrew, B. C , northwards. The other related species, A. tenuifolia. is in more sheltered waters than A. marginata. and is characteristically associated with strong t i d a l currents when i t is found in sheltered conditions.  The most extensive  habitat of A. tenuifolia is in the eastward v i c i n i t y of Vancouver Island, where there are two boundaries between A. marginata and. A., tenuifolia. at the northern end of Vancouver Island, in Queen Charlotte Strait, and at the southern end, in Juan de Fuca Strait. signs of intermediate populations.  In Queen Charlotte Strait there are no The two species meet in a sharp boundary  near the Klucksiwi River, on the northeast shore of Vancouver Island.  In  Juan de Fuca Strait, there are a long series of intermediate populations on the more exposed shores from Race Rocks, through Victoria, along the southwest shore of San Juan Island, to the west shore of Whldbey Island. from these intermediate populations that A. valida and A. valida f. have been described.  It is longipes  The latter form, having a long stipe, is probably closer  to 4 . tenulfoj-ia than the former, but both forms more properly belong to  - 6l 4. marginata. A. marginata forms the southern limit of Alaria in California. The localities of the stations from which the oceanographic data shown in Figure 6 were taken are shown in Figure 24.  There does not appear to be  any feature of the physical environment which would account for the multitude of intermediate populations in Juan de Fuca Strait and the lack of them in  :  Queen Charlotte Strait, at the boundaries between A. marginata and A. tenuifolia .  However, the inshore waters at Race Rocks, which support one of the  intermediate populations in Juan de Fuca Strait; and at Pulteney Point, which is close to the boundary at Klucksiwi R„, in Queen Charlotte Strait; both have the same long term mean salinity, 31.1%©(Anon., 1957, 1958, 1959). This coincidence may not be significant, since some of the stations to the west In A. marginata habitat, such as Umatilla Reef, Washington, have mean salinities of less than 31-  So, i f this ischaline has any significance  in the relative distributions of A. marginata and A. tenuifolia. i t must be that some factor associated with the higher salinites acts as a barrier preventing A . tenuifolia from spreading westwards.  Possibly supporting this sug-  gestion is the fact that A. tenuifolia is absent from a l l the inlets on the west coast of Vancouver Island, even from situations with strong t i d a l currents and relatively low salinity, such as Matlsett Harrows and Quatsino Narrows. Representative specimens examined: SOUTHEAST ALASKA: Sitka, June, I910, A. valida. Gardner 22^1 (UC 395389); Yakutat Bay, June, 1899, A. cordata, Saunders 230 (FH). BRITISH COLUMBIA: Table I . , June 10, 1937, A. valida,, and R.B. McCabe '4008-A (UC 633999); Sooke, June 30, 1930, A. valida. Setchell and Parkes (UC.463988); Klucksiwi R., July Ik, 1946, A. valida., Scagel 125 (UBC 205);~Deer Island Reef, 27/6/53, 4. valida. Scagel 6l6 (UBC 1993); Hope I . , 11/8/53, Scagel 952 (UBC 2167). WASHINGTON: West coast of Whidbey I . , A„ valida,. Gardner 11 (UC 9^3) > Neah Bay, June, 1917, A. valida, Gardner 3855 (UC 395392). OREGON: Brookings, 23/5A4, Doty 4057 (FH). CALIFORNIA: Pescadero P t . , Carmel Bay, 17 July, 1943, Papenfuss (UC 918077); Crescent City, June 22, 1915, H.E, Parks (uc 292573).  - 62 Alaria nana  Schrader Minn. Bot. Stud, 1:157. 1903.  Synonyms:  Alaria -praelonga f. nana (Schrader) Setchell, in Collins, Holden, efc Setchell, 1905, Ho. 1292. Alaria marginata f. nana (Schrader) Collins, 1913, p. 110.  Illustration:  Figure 39.  Notes on the type: The type locality was stated to be Port Renfrew, B. C , and, judging from Plate 23 of Schrader, is at Botany Beach In that v i c i n i t y . authentic herbarium material was found in the herbaria searched.  No  The largest  individual photographed In Plate Zk of Schrader"s paper is here designated the leetotype.  This illustration is reproduced in Figure 53.  Geographical distribution: Alaria nana Is found in the mid-intertidal zone, in areas most exposed to surf, along the coast from Yafcutat, Alaska, to California.  The  distribution is illustrated in Figure 25. Relation to other species agd the environment: A. nana is often found closely associated with Postelslaj, but not always so.  Sometimes Postelsia may be found growing without A. nana (as at  Perez Rocks, B. C . ) , and sometimes A . nana, is without Postelsia (as at Kains I . , B. C . ) , in a pattern which indicates that the ecological requirements of the two taxa are similar but not identical.  There is also a series of popu-  lations of A . nana which grows in relatively sheltered waters in Queen Charlotte Strait and Johnstone Strait,-.B;C. These populations occupy a mld-intert i d a l zone where A . tenuifolia is the species close to low water mark, and are not parallel to any of the populations found in Juan de Fuca Strait. The relationship between A . nana and A . marginata is discussed on page 60.  - 63 Repre sentative specimens examined: BRITISH COLUMBIA: Amphitrite Point, A. lanceolata, Macoon 67 (UC 160743); Plover I . , nr. Hope I . , 26/6/53, Scagel 569. (UBG 201); Deer I„.Reef, 27/6/53, Scagel 606 (UBC 187). CALIFORNIA: Land's End, Nov., 1902, A. praelonga f. nana. Setchell and Gardner,(BM).  6k  Alaria ochotensis Synonyms:  Yendo Mon„ Gen„ Alaria Ok „ 1919. none.  Illustration:  Figure ko.  Notes on the type: The type locality was stated to be the southern part of Saghalien. The type specimen (Tl) was designated by Yendo himself. on the 22 July, 1906, at Dobuki, Saghalien.  It was collected  It is illustrated in Figure 54.  Geographical distribution: The distribution of A.. ochotensjLs in Saghallen and Kamchatka appears discontinuous, but is probably actually continuous in populations around the inner shores of the Okhotsk Sea.  Lack of material from the  U.S.S.E. prevented clarification of this point.  The distribution is i l l u s -  trated in Figure 26. Relation to environment: The original description stated that the stipe is adorned by two ranks of persistent sporophyll scars.  In some specimens, however, some at  least of these projections are actually short haptera.  The phenomenon of  haptera adorning the stipe of sheltered water forms has been noted i n another genus, laminaria (L. saccharina f. linearis  J . A g . , 1867, p. 12, quoted in  Setchell and Gardner, 1925, p. 596.). Representative specimens examined: KAMCHATKA: Petropavlovsk, Aug. 10, 1930, B. Umeno (Fac, A g . , Sap.). SAGHALIEN: Shiranushi, Aug. 7, 1906, Miyabe (Fac. A g . , Sap.); Airoppu, Aug. 19, 1927, Tokida kjX (UG 360326); Kaihyoot00, 7/1930, Tokida 466 (Fac. Ag., Sap.); Kaiba-to-Kotan, 26/7/1930, Tokida (Fac. A g . , Sap.); Pilevo, Aug. 13, 1906, Miyabe (Fac. A g . , Sap.); Nayoshi, Aug. 12, 1906, Miyabe (Fac. ! A g . , Sap.); Amobetsu, West Coast, Aug. 12, I906, Miyabe (Fac. A g . , Sap.); Ochopoka, July 31, 1906, Miyabe (Fac. A g . , Sap.).  -  Alaria paradises Synonyms:  65 -  (Miyabe et Nagai) nov. comb. Pleuropterum paradlseum Miyabe et Nagai, 1 9 3 2 , p. 127• Pleuropterum paradlseum f. brevlpes Miyabe et Nagai, 1 9 3 3 , P. 101. Pleuropterum fasciculatum Yamada, 1 9 3 5 , P» 1 6 . Pleuropterum paradlseum f „ fasciculatum (Yamada) Nagai, 19^0,  Illustration;  p.  118.  Figure kl.  Notes oil the type: The localities were stated to be SJekinezakl and Minamiura on Araito I .  The type specimen (Fac. A g . , Sap.) was annotated as such by the  authors themselves and corresponds to their illustration. at Sekinezaki by H. Ito and G. Kumori in August, 1 9 2 6 .  It was collected  It is illustrated in  Figure 5 5 . Unfortunately, the plant from Minamiura was also labelled as a co-type, so the Sekinezaki plant must be designated here as a lectotype. Geographical distribution: A p a r a d i s e a has been found in the North, and Middle Kurile Islands. The distribution is illustrated in Figure 2 7 . Relation to other species: The reasons for including this taxon In the genus Alaria are discussed on page kl. Representative specimens examined: NORTH KURILE ISLANDS: Sekinezaki, Alaid I . , Aug. 1 0 , 1 9 2 6 , P. paradlseum. E. Ito and S. Kumori (Fac. Ag., Sap.); Minamiura, Alaid I , , Aug. 1 , 1 9 3 0 , £ . paradiseum, Nagai (Fac. Ag., Sap.); Nemo-wan, Onneketan I . , Aug. 1 5 , 1 9 3 5 , P. paradlseum. Nagai (Fac. Ag., Sap.). MIDDLE KURILE ISLANDS: Kitajima, Usshir I ; , Aug. 1 2 , 1 9 3 5 , Z« paradlseum, Nagai (Fac. Ag., Sap.); Minamijima, Usshir T . , Aug. 1 5 , 1 9 3 5 , P. paradisetim, Nagai (Fac. Ag., Sap.); Broughton Bay, Shimushir I . , July 2 2 , 1 9 3 0 , P. paradiseum f. breyipes. Nagai (Fac. A g . , Sap.); Yema, Urup I . , VIIl/1933, £ . fasciculatum, Yapiada (SAP 15356).  - 66 Alaria praelonga Synonyms:  Kjellman Svensk Vetens„~Akad.' Hand. 23J38» 1889. Alaria corrugata Miyabe, 1902, p. 55Alaria yezpensis Miyabe, 1902, p. 54.  Illustration;  Figure 42.  Notes on the type; The type locality was stated to be Bering Island (Commander Is., U.S.S.R.).  Nine sheets from KjeLlman's collections, one at Lund (LD 2077),  one at Stockholm (S), and seven at Uppsala (UPSV), are labelled in his hand as A. praelonga from Bering Island.  The specimen at Lund (LD 2077), collected  14-19/8/1879, is here designated the leetotype.  It Is illustrated in Figure  56. Geographical distribution: Alaria praelonga is found close to low water mark in moderately exposed areas in the Aleutian, Commander, and Kurile Islands, and northeast Hokkaido.  The distribution is Illustrated in Figure 28.  Relation to other species: 4. praelonga and 4 . marginata are very closely related species. As has already been noted (p. 39), the distance function analysis showed that the two species can be distinguished on a morphological basis.  An-  other reason for keeping the two species separate, at least for the time being, is that A. praelonga appears to have an ambiguous relationship to 4 ' crassifolia (p. 48) while 4 . marginata has an ambiguous relationship to 4 . tenuifolia (p. 60). A - crassifolia and A. tenuifolia show no signs of being related to one another, so i t would seem to be incorrect to imply this by making them closely related to the same species. Representative specimens examined: ALEUTIAN ISLANDS: Unga, Aug. 18, 1899, A. valida, cast ashore, Lawspn 5059 (UC 96673). COMMANDER ISLANDS: Bering Island, 14-I9/8/1879, K.1ellman (ID 2077). NORTH KURILE ISLANDS: Suribachi-wan, Paramushir I . , Aug. 4, 1932, A. dollchorachis. Nagaj (Fac. Ag., Sap.); Murikami-wan, Paramushir I . ,  6? July 27, 1930, 4 . macroptera. Nagai (Fac« Ag„, Sap„); Kataoka-wan, Shumushu I . , July 28, 1930, A. macroptera,, Hagai, (Fac, Ag., Sap.); Nemo-wan, Onnekotan I . , Aug. 15, 1935, Nagai (Fac. A g . , Sap.). MIDDLE KURILE ISLANDS: Broughton Bay, Shimushir I . , 22/7/1930, 4 . tenuifolia f.. typica. Nagai (SAP 026539); Yema, Urup I . , VIII/.I.93I, A.macroptera,, Yamada (SAP G26506); Yamatowan, Matsuwa I . , Aug. 14, 1935, A„ angusta. Nagai (Fac. Ag.,,Sap.). SOUTH' KURILE ISLANDS: Moyoro, Etorofu I . , Aug. 13, 1931, A. roacroptera,, Nagai (Fac. Ag., Sap.); Tosnirari, Etorofu I . , Aug. 7, 1931, A„ angusta. Nagai (Fac. Ag., Sap.); Nalbo, Etorofu I . , Aug. 27, 1934, A. macroptera, Nagamine (Fac. A g . , Sap.); Cape Atoiya, Kunashiri I . , Aug. 1, 1929, A. macroptera. Nagai and Shunamura (Fac. A g . , Sap.); Chinomichi, Kunashiri I . , Aug. 28, 1925, 4 . macroptera Tokida (Fac. A g . , Sap.); Shakotan, Shikotan I . , July 28, 1934, Nagai (Fac, Ag., Sap.). f  - 68 (Bory) Greville Alg. B r i t , xxxix. I83O.  Alaria p y l a i i  *  Synonyms,:  Agarum, p y l a i i Bory, 1826, p. ±9k Laminaria p y l a i i (Bory) De l a Pylaie, 1829, P- 31» Orgya p y l a i i (Bory) Trev., 1845, p. 28. '  Illustration;  a  Figure 4-3.  Motes, on the type: The type locality was stated to be Newfoundland, and the specimen was collected by De l a Pylaie.  There are two sheets in the Montagne Her-  barium (PC), stamped "Her'bier Bory de St. Vincent" with "De l a Pylaie" added in writing.  The two sheets are parts of the same plant.  One sheet is anno-  tated : "Partie mitoyenne de l a fronde du Laminaria P y l a i i . La partie sunerieure etant toute^deterioree. De St. Pierre et Miquelon, raportee par Lapylaie donnee en 1824." y  and the other: "J'ai rapporte cette espece de l a rade de L ' i l e St. Pierre et Miquelon pres de 1817. Terre neuve (Amerique Septle.) De l a Pylaie." There are a number of other fragmentary specimens at Paris and Lund, so these two sheets are here designated the lectotype.  The basal portion of  the plant is illustrated in Figure 57 • Geographical distribution: Alaria p y l a i i has been collected in Newfoundland, Nova Scotia, and the North Kurile Islands.  The distribution is illustrated in Figure 29.  The erroneous nature of the entity, which has been identified in the past by Scandinavian workers as A . p y l a l j , is discussed on page 52. At the present stage of knowledge, this species has a very restricted and disjoint distribution, in the Canadian Maritime Provinces and the Northem Kurile Islands.  However, the sporophyll form appears to be distinctive,  and there are sufficient  collections extant (three in the Maritimes and two  69 in the Kuriles) to give this form status as a taxonomic feature rather than dismissal as an abnormality.  Also, sosae other collections in both areas  show tendencies towards A. p y l a i i . without being clearly.identifiable as such.  Newfoundland (Canada) and the Northern Kurile Islands in some ways  have a very similar environment, having much fog and being at the limit of winter ice.  It is possible that there is some factor in the physical environ-  ment held in common by these two areas, which is connected with the distinctive morphology of this species. Representative specimens examined: MARITIME PROVINCES: St.. Pierre and Miquelon, 18.17, Agarum p y l a i i . De l a Pylaie (PC); Digby, N. S., July 2-7, 1901, Howe (NY). LABRADOR: North shore Newfoundland, 1957, 581 (Wilce). NORTH KURILE ISLANDS: Harunukotan I . , Aug. 16, 1935 (Fac. Ag., Sap.); Odomari, Onnekotan I . , Aug. 15, 1935, A. tenuifolia f. typica. Nagai (Fac. Ag., Sap.).  - 70 -  Alaria taeniata Synonyms;  Kjellman Svensk Vetens„-Akad. Hand.  23.236.  1889.  none,  Illustration;  Figure 4 4 .  Notes on the. tyjgej The type locality was stated to be Bering Island, Commander Is., U.S.S.R.[(drift).  There is one specimen at Uppsala (UPSV) labelled 1Alaria  taeniata Kjellm. (mscr.) Beringshaf, BeringBn 1 8 ^-gffi in his hand.  Since i t was stated that only one specimen was collected, this  must be the holotype. .';  7 9 p„ R, Kjellman"  It is Illustrated in Figure. 5 8 .  Geographical distribution; Alaria taeniata Is the only species, other than A. fistulosa. com-  monly found in the Gulf of Alaska area.  The distribution is illustrated in  Figure 3 0 . Relation to environment; A. taen,lata grows in many situations in the Gulf of Alaska, from the most surf-exposed to moderately sheltered l o c a l i t i e s , from the subtidal to the mid-intertidal zone.  The distance function analysis showed that samples  from Adak I. (in the Aleutian Is.), and Sitka (in southeast Alaska) also belong to this species.  These samples much extend the range.  In both l o -  c a l i t i e s , the species is uncommon, and was found growing in extremely sheltered areas. Representative specimens examined: COMMANDER ISLANDS: Bering I . , 1 4 - 1 9 / 8 / 1 8 7 9 , Kjellman (UFSV). GULF OF ALASKA: Kukak Bay, 4 / 7 / 1 8 9 9 , A. laticosta, Saunders 331 (FH); Karluk, A. p y l a i i . Aug. 20, I 8 9 9 , Setchell 5073 (UC 9 6 6 4 5 ) ; Uyak Bay, Aug. 2 0 - 2 3 , 1 8 9 9 , A. praelonga. Setchell and Lawson 5130 (NY)..  1  - 71 Alaria tenuifolia Synonyms:  Setchell, i n Collins, Holden, et Setchell Phyc. Bor. Am. XLV. 1901. " Alaria f r a g i l i s Saunders, 1901b, p. 42,5. Alaria fragilis f. bullata Saunders, 1901b, p. 425. Alarja tep,uifp,lia f. amplior Setchell, in Setchell et Gardner, 1903, p. 2j4.  Illustration:  Figure 45.  Notes on. the tyjge: The type locality was stated to be the west shore of Amaknak I . , Unalaska.  The type specimen was designated by Setchell himself (UC 96659).  It is illustrated in Figure 59. Geographical distribution: Alaria tenuifolia occurs commonly In the waters east of Vancouver Island; B. C , and more occasionally in sheltered localities in Southern Alaska and the Aleutian Islands.  The distribution is illustrated in Figure  31. Relation to other species and environment1 Alaria tenuifolia is usually found in the subtidal zone up to low water mark, in extremely sheltered l o c a l i t i e s , and is frequently associated with strong t i d a l currents.  In its major habitat in the Strait of Georgia  area, i t Is an annual, and is adapted to high seawater temperatures in summer by being fertile only in winter. The relationship between A. tenuifolia and A. marginata in B r i tish Columbia is discussed under A. marginata (p. 60). Outside British Columbia, the species is very scattered in its distribution.  The two records from the Aleutian Islands, from the type locality  and the writer's collection from Chicagof  Bay, Attu, are from populations  of f a i r size and definitely have a separate identity from the other species in the area.  The distance function analysis shows that the Attu collection  - 72 belongs with A, tenuifolia.  With the scattered Atlantic records (Appendix  III, Table ;l), the possibility exists that these are merely a few individuals showing an extreme ecological adaptation to the sheltered water habitat, and do not represent discrete populations.  However, these individuals  are morphologically indistinguishable from A. t e n u i f o l i a „ and there seems no valid reason at present for excluding them from the species.  However, this  extension of the range of 4 . tenuifolia into the Atlantic is only tentative, since i t is not supported either by any personal observations by the writer or by any statistical evidence. Representative specimens examined: GREENLAND: Sukkertoppen, A. despreauxli. Berggren, (LD 2088). ALEUTIAN ISLANDS: Amaknak I . , June 4, 1899, Setchell. 3286 (UC.96659). GULF OF ALASKA: Orca, ^Aug. 25-27, 1899, 4 . p y l a i i . Setchell and Lawson 51.53 (UC 96646). SOUTHEAST ALASKA: Douglas, Eldred ieune (Gardner 646. UC 96660). BRITISH COLUMBIA: Esquimalt, Mar., I859, A. p y l a i i . S..B. Wood (K); Oak Bay, June, 1917, Gardner 3903 (UC 395393); WASHINGTON: Roche Hbr,;-, July l 6 , 1901, Gardner (UC 96656);.Friday Hbr., July, 1917, Gardner 2308 (UC 395384); Waldron I . , July, 1910, Gardner 2250 (UC 395391).  - 73 DISCUSSION AID CONCLUSIONS Evaluation of techniques One of the greatest difficulties of a systematic study of the genus Alaria is the frequent occurrence of exceptions to the generalisations made at intermediate stages in this study.  Three of the southern limits for the  genus agree well with the 16 C isotherm of maximum sea temperature, but in Japan, Alaria crassifolia can grow in water as warm as 24*C»  Species i n the  Pacific Ocean show a measurable degree of geographical differentiation, but no such differentiation is apparent in the temperate Atlantic Ocean. Three of the species studied carefully--A. asojilenta,, A. marginata,, and A„ nana—> are approximately biennials, but A. tenuifolia is an annual.  In three of  the species—4" marginata, A. nana, and 4» tenaifolia—-no lateral growth of the blade from the midrib could be found, but this is a very noticeable feature of A. esculenta.  In another three of these species—-4» esculenta.  4. marginata. and A. tenuifolia—no measurable growth occurs in the trunk away from the transition zone, but such growth does occur in 4° nana. Moreover, i t is not even certain that these observations are general for the species named. The frequent occurrence of exceptions to almost any generalization is also reflected in the difficulties encountered in the statistical analysis, which were discussed earlier (p. 28).  The key difficulty, heterogeneity of  covariance, endangers the mathematical assumptions on which the discriminant-and-distance function analysis is at present based.  The analysis was  used, nevertheless, as i t s t i l l appeared to be the best available, and because the heterogeneity of covariance i t s e l f appeared to be a useful object of study.  The use of a s t a t i s t i c a l technique in a situation where the mathematical assumptions are not necessarily valid requires caution.  In this  case, duplicate subsamples were used as an external check on the results produced with the analysis (p. 30).  If subsamples chosen from the same  population are shown to be very similar i n the results of the analysis, i t is reasonable to accept the reality of results which show other populations to be similar to one another.  So, the tendency found in the results for  samples to be arranged in series showing a complete range of form from one population to another is also shown to be real (p„ 30). After i t has been demonstrated that the results obtained are real, there are perhaps three questions which should be asked i n evaluating the techniques used in the solution of any problem*  F i r s t , how can the tech-  niques be improved? Second, in their improved form, were these techniques sufficiently successful that they would bs used again i f the work were to be repeated?  Third, what is the relevance of such techniques to future  work? The techniques could most readily be improved in any specific case by improving the choice of the characteristics used, that i s , by rejecting the "poor" characteristics and replacing them by "better" ones, i f necessary. The computer program used provided the data needed for such decisions giving an estimate of the contribution made by each characteristic to discrimination. This estimate showed that the characters ranged from "fairly good" to "fairly poor" and that there was no clear distinction between "good" and "bad" characters.  The fact that a i l ten eigenvalues in the f i r s t run were  significant (p. 33) indicated that the use of more characteristics might be needed.  In this study, there did not appear to be much prospect of improving  the technique by making i t more elegant.  Increase in precision of the  results could best be achieved by increasing the amount of data used, and this was done in Run h up to the practical limits imposed by the resources available. There does not appear to be any other technique which would have obtained more information, or the same amount of information more efficiently, i f this work were to be repeated.  A more clear-cut classification would  have been desirable, but the classification obtained was probably as good as the nature of the data permitted. The techniques used i n the present study were-found to be useful, and should be useful for any similar preliminary survey.  On the other hand, i t  would probably not be useful, to refine these s t a t i s t i c a l methods for subsequent, more detailed work on Alaria,.  It would probably be more efficient  to change to the use of other methods more appropriate for detailed problems with a smaller 6cope. Possible causes of variation The arbitrary groupings of populations, separated on morphological and geographical c r i t e r i a in this study, are named species.  The classic botani-  cal concept of a species is of a group of populations, self-reproducing and morphologically distinguishable, which maintain gene flow between each other, but which are reproductively isolated from other comparable groupings of population's.  However, the genera among the algae, some species of which  have been investigated to this level (e.g. S i v a , b y F^ya, Burroughs and Lodge,  1951),  form only a small minority.  1955J  Fucus. by  In most cases usage  and a superficial study b y a competent taxonomist are the only bases for maintaining a species. In this study, various populations are demonstrated to be markedly heterogeneous in growth and ecology, in a manner roughly parallel to the  morphological differences between populations. of  Any further investigations  such heterogeneity require a system of classification which is workable,  consistent, and unambiguous.  It was the purpose of this study to provide  such a classification for the genus Alaria. including the use of the word "species" without prejudice to the results of future investigations.. I n much of the discussion below, i t is mors important that the  classification  should be consistent and meaningful, than that the entities involved should be precisely defined as to rank* Once the morphological variants of the genus, from locality to locality and  from habitat to habitat, have been shown to be consistent and meaning-  ful  and have tentatively been named species, the question remains whether  the variations are primarily the result of genetical or environmental 'factors .  The demonstration and proof of the relative importance and the re-  lationships between the factors involved can only be determined experimentally.  Field studies such as the present one can make only a prima facie  case in this connection, but are s t i l l useful in limiting the possibilities and  in suggesting hypotheses for further research. In practice, an environmental effect on morphology can be described  as the result of a simple and direct action of factors in the environment upon the plant, while the genotype of the plant determines only the pattern of  growth involved.  A genetical effect on morphology can be described as  one where the form of the plant is determined directly by its genotype, while the environment provides only nutrients and other such requirements for  growth. The  width of the blade in Alaria esculenta is close to being an environ-  mental effect as defined above, since i t is quickly affected in mature plants by transplanting (p. 52).  There are indications that the fasciculate  " 77 arrangement and narrow width of the sporophylls in A. angusta are the result of splitting by the force of the surf.  If this is so, i t would be another  largely environmental effect on morphology. No purely genetical effect on morphology could be proved, but one was indicated in the relationship between the three species found around Vancouver Island.  The absence of A. tenuifolia from a l l the sheltered waters  in the inlets on the west coast of Vancouver Island, adjacent to the habitat of A. marginata. indicates that A. tenuifolia In the Strait of Georgia is much more than a sheltered water form of A. marginata.  The presence of A.  nana in the mid-intertidal zone in Johmstone.Strait) where A. tenuifolia occupies the low intertidal zone, indicates that A. nana is not a mid-intert i d a l form of A. marginata. but may be the expression of a different genotype . A more general estimate of the relative importance of genotype and environment can be made from a conisideration of the heterogeneity of covariance found in the genus (p. 28; Appendix II, Table 3).  Covariance is  a measure of the degree to which the variation of a certain characteristic . in a population obtains from the variation of other related characteristics. The covariance matrix of morphological characteristics in a number of populations  is a measure of the way in which the form of the individuals in  these populations changes in response to the various circumstances encountered among the populations.  It is almost a truism to say that similar  populations have similar covariances. Variation in organisms usually results from the effects of differences in genotype or differences in environment. species of Alaria are heterogeneous.  The cov'ariances of the several  There are indications that some spe-  cies may be interfertile in certain circumstances (p.83), so the dissimilarities  - 78 in covariance can scarcely obtain from dissimilarities in genotype.  There-  fore the dissimilarities are probably the result of different environments, which would imply that differences in environment cause most of the morphological variation in the genus. Gene flow between intertidal populations is probably restricted by the fact that they are essentially in a one-dimensional array on a geographical scale, rather than in the two- or three-dimensional arrays of most populations (Bolin, 19^9).  Alaria fistulosa. which is unique among the species  of Alaria in its floating habit, is found in drift hundreds of miles from where i t grows.  Its greater powers of dispersal should result in a greater  gene flow than is the case with other species.  If variance within the spe-  cies is the result of genetic heterogeneity, A. fistulosa should have one of the lowest variances.  Reference to Appendix II, Table 5 , shows that this  species has one of the highest variances.  This would indicate that the  genetic component of variance is relatively small. Most factors in the physical environment have a marked seasonal periodicity.  Much of the tissue of an Alaria plant is continuously renewed, so  that many features (such as the lamina and sporophylls) w i l l usually have been formed in the same season as the plant is collected.  The study of  duplicate samples taken from the same population in different seasons is probably a close approach to an experiment testing the effects of different environments on the morphology of plants having a similar genetic constitution.  Two other factors which might also cause differences in morphology  are controlled to some extent by the s t a t i s t i c a l methods in the present study.  The use of a logarithmic tiansformation of the data minimizes the  effect of simple size differences, and the criterion of using only mature plants minimizes developmental differences.  The relatively large  differences found between duplicate samples taken at different seasons.(Appendix II, Table 2) again indicate the relatively large role of environment in morphological variation. Effects of environment on. distribution In this study, only the four southern limits of the genus, and the distribution of the three species of Alaria occurring in Vancouver Island waters, were considered to be known in sufficient detail for a meaningful study of the relation between distribution and oceanographic conditons. tional limits were chosen for special attention.  Distribu-  Distributions are usually  better known near their limits, since the taxon Involved is usually given more attention by collectors.  On a p r i o r i grounds, i t is only at a d i s t r i -  butional limit that i t is reasonable to expect that a taxon would be in c r i t i c a l relationship to some feature of the environment. Much of the pioneer work suggesting the use of organisms as indicators of oceanographic conditions has involved the plankton (Russell, 1935), and may not be completely relevant to similar studies of benthonic organisms. Since benthonic organisms remain fixed in one place during most of their l i f e history, they may experience a much greater range of oceanographic and other conditions during their l i f e than do plankton, which typically remain in the same water body and move with i t .  Some condition existing in the past may  have been of more significance to a benthonic organism than are the conditions observed coincidently with the presence of the organism.  In studying  the distribution of benthonic organisms, i t is more appropriate to plot the variation of the relevant conditons with respect to time at certain specific l o c a l i t i e s , than i t is to correlate the general distribution of the organisms with a synoptic survey of conditions at one particular time. An examination of Figures 6 to 9 (in which two years of data from  - 80 localities just inside and just outside the southern limits of the genus are compared) shows that i t is unlikely that any relation between Alaria and its environment w i l l be consistent for a l l species of the genus.  Sea  temperatures above 16° C appear to be limiting in the English Channel, both towards the south and towards shelter from surf.  In New England and Califor-  nia, i t might be deduced that increased sea temperature is limiting towards the south and decreased salinity is limiting towards shelter from surf.  How-  ever, Ajarla i n Japan exists where temperatures are higher, and salinities are lower, than those which might be limiting in New England and California. It appears possible that limiting factors are those relatively high or low for the area concerned.  It is possible that a. limiting low salinity, i f i t  exists, is lower in California than i t is in New England, since in general salinities are lower in the Pacific than in the Atlantic Ocean.  This is an  indication that the evolution of present ecological preferences may have occurred recently, may be of restricted geographical extent, and may be subject to rapid change in terms of geological time. A sea temperature of 16 C must have an important effect upon the physiology of Alaria. since a l l species appear to be either limited by i t , or are adapted to become fertile only when the sea temperature is below l6°C. The presence of winter sea ice gives a ready explanation for the greater depth of the lower limit of Alaria in the Arctic.  Ice grinds away  plants in the intertidal and upper subtidal zones, giving a selective advantage to those plants which can become adapted to growing at greater depths. Possible course of evolution A final consideration obtains from the s t a t i s t i c a l analysis.  It is  the apparent absence of geographical differentiation in the temperate Atlantic and presence in the temperate Pacific Ocean.  This pattern i n Alaria  - 81 Is paralleled In the algal flora in general.  In the writer's experience,  a phycologist would have no difficulty in distinguishing between a shore as being in Japan, the Aleutian Islands, the Gulf of Alaska, or the North American Pacific south of Yakutat.  He would have great difficulty in d i s t i n -  guishing between a shore in the Canadian Maritime Provinces, Iceland, the British Isles, or Norway.  The latter group of shores is more isolated geo-  graphically one from another than the former.  For instance, Iceland is more  isolated from the British Isles than the Gulf of Alaska area is isolated from Southeast Alaska.  Some presence-and-absence factors (Figure 5) might tend  to make the difference in the amount of genetic isolation less than the difference in geographical isolation.  Sandy shores in the Pacific may iso-  late rocky shores which occur in the same coast to a greater degree than a similar distance of open water, since Alaria plants may be transported alive across open sea, but may be washed up and die on sandy shores.  Strong ocean  currents in the Atlantic may reduce the isolation of rocky shores there by facilitating the dispersal of l i v i n g Alaria plants.  However, i t is impro-  bable that these two factors nullify the effects of geographical isolation to such an extent as to produce a greater degree of geographical differentiation in the Pacific. Evolution of more species i n the temperate Atlantic than the one which commonly occurs there may be restricted by an absence of suitable habitat just outside the present area of distribution for A. esculenta.  There are  large stretches of sandy beach immediately to the south of the southern limit of 4 . esculenta in New England and Brittany.  While the l6°C maximum summer  temperature may be the limiting factor here, i t would scarcely be possible for a form adapted to warmer water (such as A. crassifolia) to evolve, because of the lack of a suitable rocky shore habitat.  In British Columbia  -82  -  A., tenuifolia can grow in very sheltered waters which may be as warm as l8°C.  In these cases, the species is associated with rapid t i d a l currents.  Such habitats are in general lacking in the Atlantic, or at least, in Europe. Warm sheltered coasts, with low salinity, are usually sandy and rarely are washed by rapid t i d a l currents.  Thus i t is difficult to see how A . tenuis < •  f o l i a , also, could have evolved in Europe.  However, these considerations  only explain the non-appearance in the Atlantic, at. most, of equivalents of 4- crassifolia and A. tenuifolia.  Elimination of two species (4.  crassi-  f o l i a and 4- tenuifolia). together with A. esculenta. A. p y l a i i . and 4. grandifolia. which have a distribution in both the Atlantic and Pacific, s t i l l leaves nine species endemic to the Pacific,  line species out of four-  teen is a sufficiently large proportion to indicate that some dynamic process of speciation has been more effective in the Pacific than in the Atlantic. Geographical differentiation existing in Alaria could be explained as follows.  The genus could have arisen in the Aleutian and Kurile Island  chains, which present a great variety of semi-isolated populations and a situation favourable to the evolution of species (Sewall Wright, Indeed, the majority of present day species occur there.  1931).  Upon the disap-  pearance of the Bering Strait land bridge, the Arctic and Atlantic Oceans could have been quickly invaded by 4. esculenta and, to a lesser degree, by 4. n y l a i i , A. grandifolia. and perhaps 4. tenuifolia.  These species, occupy-  ing the upper subtidal zone in the more rocky western Bering Sea, would be in a favourable position to do this. they were closest to Bering Strait.  Being in the western Bering Sea, Subtidal species would have a better  chance than intertidal ones of invading shores exposed to the action of ice. The temperate Atlantic populations of 4. esculenta are s t i l l relatively  -83 homogeneous, but are beginning to evolve in similar directions as the species in the Pacific.  Thus some individuals in sheltered water are similar  to A. tenuifolia. some populations in-the warmer Norwegian fjords grow thick, dark, sterile sporophylls similar to those of A. crassifolia. and some populations growing high on exposed coasts are approaching A - crispa and A. nana. However, because speciation has not progressed so far in the Atlantic, these various populations are ecotypes rather than species. In general, the species of Alaria outlined in this study appears to be roughly comparable to species as they are classically defined.  How-  ever, the species in Alaria have not yet evolved to the stage where they are completely distinct.  The lack of clustering in the distance function analy-  sis indicates that there is a complete range of intermediate forms between similar species. Two terms have been developed to describe a geographical range in form and genotype within a species, the cline and the rassenkreis (Mayr, 1 9 4 2 ) . In a cline, one form merges gradually into another in one direction along a geographical range.  A rassenkreis is a cline which occupies most of the c i r -  cumference of a c i i c l e .  While each element of the rassenkrejs is inter-  f e r t i l e with the elements immediately adjacent to i t , there are sufficient differences of genotype between the two ends of the range, that the populations are intersterile i f they come in contact with one another. Examples both of a rassenkreis and a cline appear to occur in the species of Alaria found in British Columbia.  The sharp boundary between A. marginata  and A . tenuifolia in Queen Charlotte Strait, but the presence of a great number of intermediate populations in Juan de Fuca Strait, is best explained by suggesting a rassenkreis around Vancouver Island.  The two species  -8k are Interfertile in Juan de Fuca Strait, but intersterile in Queen Charlotte Strait.  In a similar way, a cline is apparent in A. marginata and 4.  nana, as they appear to be intersterile in the northern part of their range, and interfertile in the southern part. Superimposed on the problems which clines and rassenkreise present to the taxonomist are the problems which arise from the probability that environmental rather than genetic differences cause the greater proportion of variation in morphology. The general picture is of a genus where the species are s t i l l evolving rapidly.  -85BIBLIOGRAPHY Agardh, C.A. -1817. Synopsis algarum Scandinaviae, adjecta dispositione u n i v e r s a l i algarum. x l + 135 PP«> Lundae. Agardh, J . G . -1848. Species genera et ordines algarum . . . . V o l . 1. Algas Fucoideas. v i i i + 3°3 PP* Lundae. , -1867. De Laminarieis symbolas o f f e r t . Lunds Univ. A r s s . Tom. IV,  36 pp.  -I872. Bidrag t i l l Kannedomen af Gr&nlands Laminarieer och Fucaceer.  . Svensk. Vetensk.-Akad. Hand. M.F. Bd. 10 No. 8, 31 pp.  Anderson, E. -1949. Introgressive h y b r i d i z a t i o n IX + 109 PP» New York; Anon. -1952. B u l l e t i n hydrologique pour l'annee 1951. Conseil Int. pour 1'Exploration de l a Mer. Charlottenlund S l o t . -1953. B u l l e t i n hydrologique pour l'annee 1952. Conseil Int. pour 1'Exploration de l a Mer. Charlottenlund.Slot. -1954. B u l l e t i n hydrologique pour l'annee 1953* 1'Exploration de l a Mer. Charlottenlund S l o t .  Conseil Int. pour  -1957. Observations of seawater temperature and s a l i n i t y on the P a c i f i c coast of Canada. V o l . XVI, 1956. 84 p p . , 20 f i g s . F i s h . Res. Bd. Canada. Nanaimo. -1958. Observations of seawater temperature and s a l i n i t y on the P a c i f i c coast of Canada. Vol.XVII, 1957. F i s h . Res. Bd. Canada. Manuscript Report Series (Ocean, and Limn.) No. 23. 100 p p . , 17 f i g s . Pacific Oceanographic sGroup, Nanaimo. ! -1959. Observations of seawater temperature and s a l i n i t y on the P a c i f i c coast of Canada. Vol.XVIII, 1958. F i s h . Res. Bd. Canada. Manuscript Rep. Ser. (Ocean, and Limn.) No. 48. 62 p p . , 29 f i g s . P a c i f i c Oceanographic Group, Nanaimo. -1962. Surface water temperatures and s a l i n i t y . P a c i f i c coast, North and South America and P a c i f i c Ocean Islands. U.S.C. and G.S. Publ. 31-3First Edition. 0.6] + 71 p p . , 2 t a b l e s . Washington. Baardseth, E. -1956. The growth rings i n A l a r i a s t i p e s . 2nd International Seaweed Symposium, Trondheim, pp. 153-157, 1 f i g . Batters, E.A.L. -1902. A catalogue of the B r i t i s h marine algae. J . Supplement. 107 pp. London.  Bot.  -86-  Blackith, R . E . -1957• Polymorphism in some Australian locusts. 13:183-196, 2 f i g s . , 1 table.  Biometrics  - I 9 0 5 . The algae-vegetation of the Faeroese coasts with remarks on the phytogeography. pp. 44-834, p i s . x i i i - x x i v , figs. 151-164. Reprinted from, J . E . B . Warming, Botany of the Faero&s, Part II, Copenhagen. -1907» Note on the question whether Alaria esculenta sheds its lamina periodically or not. Bot. Tidsskr. 28:199=202. Bolin, R . L . - 1 9 4 9 . The linear distribution of intertidal organisms and its effect on their evolutionary potential. Xlle Congres International de Zoologie, Communications 6eme section, p. 4 5 9 . (not seen. Quoted in Doty, 1951, P.568) Bory de Saint-Vincent, J.B.G.M. 9:187-194.  - 1 8 2 6 . Laminaire. Diet. Class. Hist. Nat.  Burrows, E.M. and S. Lodge - 1 9 5 1 . Autecology and the species problem in Fucus. Jour. Mar. B i o l . Assoc. U.K. 3 0 : J.6I-I76, 6 f i g s . , 1 table. Chapman, V . J . - 1 9 4 2 . Zonation of marine algae on the sea-shore. Proc. Linn. Soc. 154:239-253. Collins, F.S. -1913. The marine algae of Vancouver Island. Cam" Geol. Survey. Victoria Mem. Mus. B u l l . No. 1 : 9 9 - 1 3 7 . , I. Holden, and W.A. Setchell. -I895-I919. Exsiccati Fascicles 1-46 and A-E. Phycotheca Boreali-Americana. Maiden, Mass. Day,  C.G. -1959* Oceanographic observations, 1 9 5 8 . East coast of the United States. ' U.S. Fish and Wildlife Service. Special Scientific Report Fisheries No. 3 1 8 . v i + 119 pp., 18 f i g s . , 18 tables. Washington. 0  - i 9 6 0 . Oceanographic observations, 1959. East coast of the United States. U.S. Fish and Wildlife Service. Special Scientific Report - Fisheries No. 3 5 9 . iv + 114 pp., 18 f i g s . , 18 tables. Washington. De La Pylaie, A.J.M.B. -1824. Quelques observations sur les productions de l ' i l e de Terre-Neuve, et sur quelques algues de la cote de France appartenant au genre Laminaire. Ann. S c i . Nat. 4:174-184. - 1 8 2 9 . Flore de l ' i l e de Terre Neuve et des lies St.Pierre et Miclon. 128 pp. Paris. Dickie, G. -1853* Notes on flowering plants and algae collected during the voyage of the "Isabel", pp. 133-144 In, E.A. Inglefield. A summer search for Sir John Franklin; with a peep into the polar basin. 232 pp., London. Dixon, P.S. - i 9 6 0 . Taxonomic and nomonclaxural notes on the Florideae, II Bot. Not. 113 ( 3 ) : 2 9 5 - 3 1 9 .  - 6 >  Doty, M. S. - I 9 4 7 . The marine algae of Oregon. Phaeophyta. Farlowia 3 ( l ) : I - 6 5 , 10 p i s .  Part I.  Chlorophyta and  -1957. Chapter i8„ Rocky i n t e r t i d a l surfaces, pp. 535-585,. 18 f i g s . , 1 p i . In, J . W. Hedgepeth, e d . Treatise on marine ecology and paleoecology. V o l . I. Ecology. Geol. Soc. of America Memoir 67. v i i i + 1296 pp. Washington. F a l l i s , A . L. - 1 9 l 6 . Sta. 1:137-155.  Growth i n some laminariaceae.  Publ. Puget Sound B i o l .  F i s h e r , R. A . -.1936. The use of multiple measurements i n taxonomic problems. Ann. Eugenics 7:179-188, 1 f i g . , 9 t a b l e s . F o s l i e , M. - I 8 8 6 .  K r i t i s k fertegnelse over Norges havsalgaer e f t e r aeldre  botaniske arbeider i n d t i l aar I.85O,  Troms^i Museums Aarshefter 9:85-137.  Fdyn, B. -1955. S p e c i f i c differences between the northern and southern European populations of the green alga Ulva lactuca L. Pubbl. Staz. Zool. Napoli 27:261-270, 1 f i g . , p i . 17. Frye, T . C. -1915. The kelp beds of southeast Alaska. Part IV, pp. 60-104. In. F. K. Cameron. Potash from kelp. U.S. Dept. of Agriculture Report No. 100 Washington. , G. B. Rigg, and W. C. Crandall - I 9 I 5 . The size of kelps on the p a c i f i c coast of North America. Bot. Gaz„ 60:473-482, 2 p i s . Ghobi, Kh. Ya. -1878a, A l g a l f l o r a of the White Sea and of the parts of the northern A r c t i c Ocean adjacent to i t . Trans. S t . Petersburg Soc. Nat. 9:293-412. ( i n Russian). __.  -1878b. Die Algenflora des Weissen Meeres und demse.lben zunachtsliegenden Theile des nBrdlichen Elsmeeres. Memoires de 1'Academie Imperiale des Sciences de S t . Petersburg. V I I serie Tome XXVI No. 1. 92 pp. e  Gmelin, S. G. - I 7 8 6 .  H i s t o r i a Fucoruia.  viii.  + v i + iv  + 239 P P » , 3^ p i s .  Goodenough, S., and T . J . Woodward - I 7 9 7 , Observations of the B r i t i s h F u e l , with paz-ticular descriptions of each species. Trans. L i n n . Soc. 3^84235, 4 p i s . Gray, S. F. -1821. A natural arrangement of B r i t i s h Plants . . . . V o l . I. x x v i i i + 824 p p . , p i s . 1-21. London. G r e v i l l e , R. K. - I 8 3 0 . London„  Algae Britannicae.  Gunnerus, J . E. - I 7 6 6 . Nidrosiae.  F l o r a Norvegica.  l x x x v i i i + 218 p p . , 19 p i s . Pars p r i o r , v i i i + 100 p p . , 3 p i .  --88--  Harvey, W.H. -i860. Characters of new algae, chiefly from Japan and adjacent regions, collected by Charles Wright in the Horth Pacific exploring expeditions under Captain John Rodgers. Proc. Amer. Acad, of Arts and Sciences 4:327-335. Horneman, J.W. -1827. Homenclatura florae danicae emendata cum indiee systematico et alphabetic©, x x y i i i + 214 pp. Eafniae. ______1837• Dansk oeconomisk Plantelaere.  [~iv] + 992 pp. Copenhagen.  Hudson, W. -1762. Flora Anglica. First Edition, xvi + 506 +  22  pp.  Isakov, I. S., et a l . , eds. Marine Atlas. Volume 2. 1953* x v i i i + 76 pis. + 5 pp. Moscow, (in Russian). Jonsson, H. -1904. The marine algae of east Greenland. 30 (l):l-73, 13 figs.  Medd. om Gr^n.  -1912. The marine algal vegetation. In, J . L . A . Kolderup Rosenvinge and J . E . B . Warming. The Botany of Iceland. Copenhagen and London. Kanda, T. -1936. On the gametophytes of some Japanese species of Laminariales. Sci. Pap. Inst. Alg. Res., Fac. S c i . , Hokkaido Imp. Univ. l(2):221-260, pis. XLVI-XLVIII, 27 figs. Kawabata, S. -I936. A l i s t of marine algae from the island of Shikotan. Sci. Pap. Inst. Alg. Res. Fac. S c i . Hokkaido Imp. Univ. 1(2):199-212. Kjellman, F.R. -1883a. The algae of the Arctic Sea. K. Svens. Vetensk.-Akad. Hand. 20(5). 35O pp., 31 pis. -1883b. Norra ishafvets algflora. B. I l l In, A.E. Nordenskiold. Vega-Expeditionens vetenskapliga Iakttagelser bearbetade af deltagare i resen andra forskare. Stockholm. Kjellman, F.R. -1889. Om Beringhafvets Algflora. K. Svensk. Vetens.-Akad. Hand. 23(8). 58 pp., 7 plates. -1890. Handbok i Skandinaviens Hafsalgflora. Stockholm.  I.  Fucoideae. [iv] + 103 PP«  -1906. Zur Kenntniss der marinen Algenflora von Jan Mayen. Archiv for Botanik 5(14). 30 pp., 3 pis. and J . V . Petersen -1887. Om Japans Laminariaceer. Vol. 4 pp. 253-280, Pis. X, XI I n , A . E . Nordenskiold, Vega-Expeditionens vetenskapliga Iakttagelser bearbetade af deltagare i resen andra forskare. Stockholm. (|  -89Lamouroux, J . V . F . -1813. Essai sur l e s genres de l a f a m i l i e des Tfealassiophytes ( l ) non a r t c u l e e s . Ann. Mus. H i s t . Nat. 20s 21-47; 115-139; 267-293. p i s . VII-XIII. Lanjouw, J . (ed.) -1961. International code of Botanical Nomenclature. 372 pp. International Bureau f o r Plant Taxonomy and Nomenclature. Utrecht. , and F.A. Stafleu -1956. Index herbariorum. Part I .  world. 3rd ed. 224 pp. Utrecht.  The Herbaria of the  L i g h t f o o t , J . -1777. F l o r a S c o t i c a . V o l . II. pp. 531-1151 + [24] pp. London. Linnaeus, C. -1767. Mantissa plantarum. Generum e d i t i o n i s VI et Specierurn e d i t i o n i s II. 142 + [2] pp. Holmiae. Lubischew. A . I . -1962. On the use of discriminant functions i n taxonomy. Biometrics 18:455-477, 9 f i g s . , 10 t a b l e s . Lund, S. -1959. The marine algae of east .Greenland. I . Taxpnomical p a r t . Medd. om Gr^n. 156 ( l ) . 247 p p . , 42 f i g s . Mahalanobis, P.C. =1936. On the generalized distance i n s t a t i s t i c s . Proc. Nat. Inst. S c i . (India) 12:49 (not seen, quoted i n Rao, 1952, p. 328) Mayr, E. -1942. Systematics and the o r i g i n of the species. XVI + 33^ PP« New York. Martens, G.M. von -1866. Die Preussisehe Expedition naoh Ost-Asien nach  amtlichen quellen. Botaniseher T h e i l . Die Tange. 152 p p . , 8 p i s . B e r l i n . , H. -I829. Zwei botanisch-wissensch&ftlicae Berichte. Linnaea 4:43-73«  Miyabe, K. -1902. Kelp gathering. Hokkaido aquatic products investigation reports V o l . 3« Hokkaido Government Colonization Resources Development Department. Aquatic Produce Bureau. 216 p p . , 4 l p i s . Sapporo, ( i n Japanese). _____[-1957]• On the Laminariaceae of Hokkaido. Jour. Sapporo A g r i c u l t u r a l College, Faculty of A g r i c u l t u r e . V o l . I. [iv] + 50 pp. (Translation of Miyabe (1902) into English by J . Tokida). Miyabe, K. -I928. On the occurrence of c e r t a i n Behring and Kurile species of Laminariaceae i n a small isolated region o f f the southern extremity of Saghalien. Proc. 3rd. Pan-Pacific S c i . Congress, Tokyo, 1926. V o l . I. B. VIII (a) 11: 954-958. and M. Nagai -1932. Pleuropterum paradlseum, a new genus and species of Alarieae from the Northern K u r i l e s . Proc. Imp. Acad. Tokyo 8(4); 127-130, 2 f i g s . and M. Nagai -1933• Laminariaceae of the Kurile Islands. Nat. H i s t . Soc. 13(2):85-102.  Trans. Sapporo  Nagai, M. - 1 9 4 Q . Marine algae of the Kurile Islands I. Jour. Fac. Agriculture Hokkaido Imp. Univ. 4 4 ( l ) . 1 3 7 pp., pis. I-III. Nash, S.W., unpublished ms. Multiple comparison methods in multivariate 9 pp. ________ and P. Jolicoeur, unpublished ms. Calculating discriminant 12 + [l] + 2 + 4 pp. N e i l l , P. - 1 8 1 5 . Fuel No.10, In/ . The Edinburgh Encyclopedia 2 3 pp.  analysis.  functions.  •  Nishibayashi, T. and S. Inoh - I 9 6 3 . On the sorus development in Alaria valida Setch. et Kjellm, A. crassifolia Kjellm., Laminaria Yendoana Miyabe, L. ciehorioides Miyabe and Eisenia bicyclis (Kjellm.) SetchT Bot. Mag. Tokyo 76': 14-23, 37 figs, (in .Japeme'se)". Nordenskiold, A . E . - 1 8 7 1 . No. 1 0 . Redogorelse foran expedition t i l l C-ronland ar 1 8 7 0 . Ofv. K. Vetens.-Akad. Forhandl., 1 8 7 0 [published 1 8 7 1 ] ,  pp.  931-1081.  Postels, A. and F . Ruprecht -1840. Illustrationes algarum . . . . v i + 2 8 + \_2~\', iv + 2 2 + [2]-, 4 0 p i s . St. Petersburg, (in parallel Latin and Russian texts). Powell, H.T. and Y.M. Chamberlain - 1 9 5 6 . Plant l i f e on Rockall. 6 pp. In, J . Fisher. Rockall. 2 0 0 pp. London.  Appendix B.  Printz, H. - 1 9 2 6 . Die Algenvegetation des Trondhelmsfjordes. Skrift. Norske Vid.-Akad. Oslo I. Matem.-Naturvid. Klasse. No. 5 . 2 7 4 pp., 1 0 p i s . , 1 map. 2 9 text figs. Rao,  C R . - 1 9 5 2 . Advanced s t a t i s t i c a l methods in biometrical research. 3 9 0 pp. New York. o  Rosenvinge, J . L . A . Kolderup -l893« Gr^nlands havalger. 3: 7 6 5 - 9 8 1 , 2 p i . ,  56 figs.  Medd. om Gr^n.  Ruprecht, F . J . »l85-l. Tange des Ochotskischen Meeres. pp. 1 9 1 - 4 3 5 , pis. IX-XVIII In, A. Th. v. Middendorff. Reise In den aussersten Norden und Osten Sibirlens. Band I . Einleitung, Klimatoiogie, Geognosie, Botanik. Theil 2 . Botanik. [vii] + 4 3 5 pp. St. Petersburg. Russell, F . S . - 1 9 3 5 . On the value of certain plankton animals as indicators of water movements in the English Channel and North Sea. Jour. Marine B i o l . Assoc. U.K. 2 0 : 3 0 9 - 3 3 1 , 6 figs. Saunders, deA. - 1 9 0 1 a . A new species of Alaria. pl. 3 3 .  Minn. Bot. Stud. 2 : 5 6 1 - 5 6 2 ,  - 1 9 0 1 b . Papers from the Harriaan Alaska Expedition. XXV. The algae of the expedition. Proc. Wash. Acad. S c i . 3:391-486, p i s . X-XXYI.  -91-  Sauvageau, M.C. -19l6. Sur l a sexualite" heterogamique d'une Laminaire (Alaria esculenta) C R . Acad. S c i . 162:840-842. Savage, S. -194-5. A catalogue of the Liaeaean Herbarium. 229 PP« London. Scagel, R . F . -1957* An annotated l i s t of the marine algae of British Columbia and Northern Washington (including keys to genera). Nat. Mus. Can. B u l l . No. 150, B i o l . Ser. No. 52. v i + 289 pp. Ottawa. _______1963. Distribution of attached marine algae in relation to oceanographic conditions in the northeast Pacific, pp. 37-50, 11 figs. In, M.J. Dunbar, ed., Marine distributions. Roy. Soc. Canada Spec. Publ. No. 5« [8] + 110 pp. University of Toronto Press. Schrader, H.F. -1903« Observations on Alaria nana sp. aov. Minn. Bot. Stud. 3:157-165, pis. 23-26. Setchell, W.A. and N.L. Gardner -1903. Algae of northwestern America. Univ. Cal. Publ. Bot. 1:165-418, pis. 17-27. _______ and -1925. The marine algae of the Pacific coast of North America, Part III Melanophyceae. Univ. Cal. Publ. Bot. 8:383-898, p i s . 34-107. Silva, P.C. -1952 A review of nomenclatural conservation in the algae from the point of view of the type method. Univ. Cal. Publ. Bot. 25. 241-324. Southward. A . J . -i960. On changes of sea temperatures in the English Channel. Jour. Marine B i o l . Assoc. U.K. 39.449-458. , and D . J . Crisp. -1954. Recent changes in the distribution of the intert i d a l barnacles Chthamalus stellatus Poli and Balanus balanoides L. in the British Isles. Jour. Anim. Ecol. 23:163-177. Stackhouse, J . 1797» Nereis Britannica Fasciculus II. Physiological observations on the structure and fructification of Fuel, pp. ix-xxiv. Bathorii et Londini. _______1809 Tentamen Maximo-cryptogamicum ordinem novum in genera et species distributum. Memoires de l a Societe Imperiale Naturalistes Moscou 2:50-97* _____  -l8l6. Nereis Britannica. x i i + 68 pp., 20 pis. Oxonii.  Stroemfelt, H.F.G. -1886. Om algenvegetationen vid Islands kuster. Akad. afhand. Vidt. F i l o . Fak. Upssala Mat.-Nat. Section Al'arosalen No. 4. 89 pp., 3 pis. Sundene, 0. -1962. The implications of transplant and culture experiments on the growth and distribution of Alaria esculenta. Nytt. Mag. for Bot. 9:155-174, 8 f i g s . , 6 pis. Svendsen, P. -1959« The algal vegetation of Spitsbergen. A survey of the marine algal flora of the outer part of Isfjorden. K. Dept. for Industri og Handwerk Norsk Polarinstitutt Skrifter Nr. 116. 49 pp., 10 tables, 2 p i s . , 10 figs.  Taylor, W.R. -1957. Marine algae of the northeastern coast of North America. 2nd ed. ix + 509 PP«, 60 pis. Univ. Michigan Press Ann Arbor. Tilden, J . E . 1894-1902. American Algae. Exsiccati. Centuries I-VI. St. Paul. Trevisan di San Leon,'V.B.A. =l845« Nomenclator algarum 80 pp. Padua.  Tom. I,  Turner, D. -1802. A synopsis of the British Fuci. V o l . 1. x l v i + 400 pp. Yarmouth. Wilce, R.T. -1959. Marine algae of the Labrador Peninsula and North-west Newfoundland (Ecology and Distribution) Nat. Mus. Canada. B u l l . No. 158 iv + 103 pp., 1 1 pis. Ottawa. Wille, J.N.F. -1897. Beitrage zur physiologiscSien Anatomle der Laminariaceen. 70 pp., 1 p i . K. Norske Frederiks Universitet Festskrift No. 4. Christiana. Wright, S. -1931 Evolution in Mendelian populations. Genetics l 6 : 9 7 - 1 5 9 > '21 figures. Yabu, H. -1957. Nuclear division in the sporangium of Alaria crassifolia Kjellm. B u l l . Fac. Fish. Hokkaido Univ. 8(3) -.185-189. Yamada, Y. -1935. Marine algae from Urup, the middle Kuriles, especially from the vicinity of lema Bay. S c i . Pap. Inst. Alg. Res., Fac. S c i . , Hokkaido Imp. Univ. l ( l ) . 26 pp., 10 f i g s . , 10 pis. Yendo, K. -I919. A monograph of the genus Alaria Jour. Col. S c i . Tokyo Imp. Univ. 43(1): 145 pp., 19 pis. Zinova, A.D. -1953. Classification of the brown algae of the northern seas of the U.S.S.R. Botanical Institute, USSR Academy of Sciences. 222 pp., 153 figs. Leningrad (in Russian). Zinova, E . S . -1929a. Algae of the Japan Sea (Browns) B u l l . Pacific Ocean Scientific Fishery Station, Vladivostok. 3(4):l-55. (in Russian). -1929b. White Sea seaweeds and their practical application. Trans. Inst. Industrial Investigations by the Arkangel Prov. Executive Committee Fasc. VI. 47 pp., 1 map. (in Russian). -1940. Marine algae of the Commander Islands. Trans.Pacific Ocean Committee V. USSR Academy of Sciences, pp. 165-243. (in Russian). m  __-1954. Algae of the Gulf of Tartary. Trans. Bot. Inst. USSR Academy of Sciences. Ser. II, Fasc. 9:311-364, 1 f i g . (in Russian).  -93Ap-pendix I - Nomenclature 1.  Discussion of. typjficatlpja, and legitimacy of names i n A l a r i a .  General There are some sources of ambiguity i n the discussion following,, which cannot be avoided.  In work by Russians, or on Russian t e r r i t o r y , before 1917,  the. Old Style Calendar may have been used.  This applies to work by Mertens  (1829), Postels and Ruprecht (l84o), Ruprecht ( l 8 5 l ) , Kjellman (1883a, I889), and Ghobi (1878a, 1878b).  Thus, Kjellman stated (1883, p. 222) that he c o l -  lected A l a r i a e l l i p t i c a i n J u l y i n the Siberian Sea, while a l l the relevant material the writer found i n herbaria i s l a b e l l e d as being c o l l e c t e d i n June. The former date i s presumably New Style Calendar; the l a t t e r , Old S t y l e . A number of descriptions were published guages.  i n p a r a l l e l texts in' two l a n -  Postels and Ruprecht (1840) used L a t i n and Russian.  of the two texts was checked, and the L a t i n was used. published apparently  The equivalence  Kjellman. (1883a, 1883b)  simultaneously separate volumes, as English and Swedish  editions; the English one was used here as the o r i g i n a l source.  1878b) was published i n equivalent Russian, and German t e x t s .  Ghobi (1878a,  The Russian  text probably appeared f i r s t , since the German contains a reference t o i t . The Russian text was used.  Miyabe (1902) was published  i n Japanese.  Vol-  ume I of the Journal of the Sappoz-o Agriculture College was reserved f o r an English e d i t i o n , which Miyabe never prepared. as a memorial to Miyabe, and published  I t was translated by Tokida  i n 1957, s t i l l as Volume I (Miyabe  1957 ). The English text was used i n t h i s study, but the Japanese was consulted f o r the i l l u s t r a t i o n s , which are somewhat d i f f e r e n t . Unless the d i v i s i o n of a species into subspecific taxa i s being d i r e c t l y discussed, the type e n t i t y i s merely referred to under the s p e c i f i c epithet;  i . e . , Alaria fistulosa is used instead of Alaria fistulosa f. fJUstulosa. The sub-specific rank was not actually used in this study,  ftoere is doubt that  some of the species studied are actually species in the classical sense.  The  sub-specific rank is reserved in case future study should show that some of the taxa currently labelled species should be reduced in rank.  A l l names which  have appeared 5.n the literature are included, even those mentioned in passing, and those which appear to have been spelling mistakes. names found only on herbarium sheets were ignored.  However, manuscript  Also, Yendo (1919)  sometimes referred to a later author's citation of a previous author's species under the name of the later author, omitting "sensu".  Thus, he referred  to "Alaria esculenta Dickie" when there was no question that Dickie (1853) was referring to A. esculenta (L.) Grev., but his identification was in some doubt.  A l l these cases were checked, but they are not included in the l i s t  below. The spelling of place names in the Kurile Islands is difficulty to regulate .  These are ad hoc Japanese transliterations of l o c a l , non-Japanese names,  transliterated again into Roman characters by one of two different systems. For instance, there is Shumushu I . (or Shimushu I.) in the North Kuriles, and Shimushiru I . (or Shimushir or Shumushir I.) in the Middle Kuriles, while Mushiru I . is yet another island.  Place names used in the present study have  been recorded as they came to hand from conflicting versions i n the literature, the reference maps used, and translations of Japanese herbarium annotations. An ill-informed standardisation could cause much confusion,  Russian versions,  such as Iturup 0 . . for Etorofu I., and Alaid 0 . for Araito I . , have been avoided although they are sometimes used by the Japanese.  There is also disagreement  about the division of the North, Middle, and South Kurile Islands.  In this  study, the lines of division are drawn through Mushi.ru Strait and Etorofu Strait.  Typifieatjon, of the, ggaus. In the most recent conservation of A^ajrla Grev„, I83O (Si.lva, 1952, p. 259), Fueus esculentus L . is cited as the type species.  Fucus esculentus L .  has at one time or another been combined under other genera, which have since been assigned quite different type species, so that Eucus^ esculentus has been excluded.  These combinations are as follows : Ceramlum esculentum (L.) Stackh., 1797, p. 24. Laminaria esculenta (L.) C. Ag., I.817, p. 13• Agarum esculentum (L.) Bory, 1826, p. 194.  Statements in the literature notwithstanding (De l a Pylaie, 1829, p. 23), the combination Laminarja esculenta was not made by Iamouroux (1813). Fupus esculentus L . is the type species of a number of generic names other than Alaria.  These names, to be rejected under Article 14, Note 3, of the  Code, occur i n the following combinations: Musaefolia esculenta (L.) Stackh., 1809, p. 53. Qrayia esculenta (L^) Stackh., I 8 l 6 , p. 8. Phasganon scoticum Walker, in N e i l l , 1815, p. 4. Phasgonon esculentum (L.) Gray, 1821, p. 385. Orgya esculenta (L.) Trey., 1845, p. 28. Podopterls esculenta (L.) De l a J y l . , 1829, p. 23. Phasganon a^atum Ruprecht. 1851, p. 86. Walker's original description of Phasganon scoticum was not published, and so is of no significance for nomenclature.  However, N e i l l , 1815, p. 4,  quoted "Dr. Walker's classification . . . . . in his own words" as a matter of academic interest.  It is a doubtful decision to reject this under Article 34(l)  of the Code as an invalid publication, since i t was not accepted by N e i l l , the publishing author. Walker in. N e i l l .  It is safer to accept this as a valid publication,  Phasganon. or "tangle", was described as follows:  "Stipitatum monophyllum, coriaceum, F r . Vesiculosae superficiales globosae, gelatinosae, inter corticem et epidermidem f o l i i tumidae."  -96-*The genus stated to include the species Ph,. baj^ifjsrjii,, or sea-belt; Eh. Mariae. or St. Mary's thistle; Ph.. esculentum  0  or common tang; £ h . scoticum.  These were equated, presumably by B e l l i , to £ . saccharlnus L„, £ . nolyschldes Lightf., £ . digitafcus L„, and F_. esculentus L„, respectively.  The four latter  species have been transferred (Lanjouw, 1961, pp. 210, 212) to Laminaria. Saccorhiza. Laminaria again., and A J a J a , which are a l l conserved genera. Thus, the typification of Phasganon Walker, i n N e i l l is an academic question. However, Gray (l82l) made Phasgojaon,, presumably sn orthographic variant, cover Alaria and Saccorhiza. and Ruprecht (1851) applied Phasganon. to Alaria only. It seems logical to typify Phasganon by the species Fucus esculentus. and so make i t another nomep re.iiclendum for Alaria. that Phasganon Walker was ever published.  Ruprecht was apparently unaware  He proposed the specific epithet  alatum on criteria of suitability and pre-Linnaean p r i o r i t y . Podopteris  {esculenta] De l a Pylaie, 1829, p. 23, involved an adequate  description of the genus, and "Lam, esculenta Lamx." was indicated as a type, although the combination was not explicitly made„  This appears to f u l f i l re-  quirements of valid publication, even though later in the same publication this generic name was ignored in favour of laminaria„ Names for. which type material found was adequate: 1.  Alaria angusta  Kjellman, 1889, p. 38, P i . III.  2.  Alaria corrugata  See page 45.  Miyabe, 1902, p. 55, PI. 24.  This species was stated to grow on the islands of Etorofu and Urup in the Kurile Islands.  The illustration is of a plant from Shana Gun on Etorofu,  so this should be the type locality of the leetotype.  There are two suitable  specimens at Hokkaido University (Fac. Ag„, Sap.), Japan.  One was collected  at Shana on August 13, I895, by Tanagawa (No. 220), with the label of "Alaria  --97corrugata Miyabe" crossed out and "dolichorachis K.1 ellm„" added. collected by S. Fujimura in August,  I89Q,  The other was  and has both "Alaria "lanceolata  Kjellm." and "macrontera" substituted, in Yendo s hand, for corrugate..  Neither  5  of these can be directly connected with Mly&be, but i t is unlikely that he had so many specimens from the Kurile Islands in 1902 thai, he would not use these. The latter specimen is here designated as the lectotype, since i t is similar to the illustration, and the annotation "macroptera," Indicates that Yendo (1919, p. 79) may have used i t in synonymicing A. corrugata with A. macroptera. 3.  Alaria crassifolia  X, figures 9 - 1 2 . 4.  Alaria erispa  5.  Alaria curtlpes  Kjellman, in Kjellman g_ Petersen, 1887, p„ 276, PI.  See page 4-7. Kjellman, 1889, p. 37, PI. I l l , figures 5 ~ 7.  See page 49.  Saunders, 1901a, p. 561, P I . XXXIII.  The type area is stated to be Monterey Bay, Carmel Bay, and Point Sur, in California.  There is one specimen in the Farlow Herbarium annotated  "Alaria _..&.§,„ Pacific Grove. Aug. 95" which also has an annotation by Kjell-' man quoted in part in Saunders' description. ders' plate.  The specimen is similar to Saun-  It appears to be a good lectotype, even though there is no men-  tion of the specific epithet on the sheet, and moreover Saunders states that o  he observed the plant in the summer of 1896. 6.  Agarum delisei  Bory, 1826, p. 194.  The type locality was stated to be Newfoundland.  The description was  based upon a single specimen obtained by Bory from Delise.  There is one spe-  cimen in the Montagne • Herbaiium at Paris (PC) labelled "Agarum delisei Bory Terre-Neuve D. Minat" in an unknown hand.  The specimen agrees very well with  the description of laurel-leaf like sporophylls well spaced along most of the length of the trunk, and therefore must be the hclotype. 7.  Alaria delisei  (Bory) Greville, I83O, p. xxxix.  --98This Is a simple transfer, explicitly made b y Greville, of Ho. 6 above, and  so w i l l have the same type.  8.  Orgya delisei  (Bory) Trevisan, 1845, p. 28.  Page 28 is simply a l i s t , but reference to p. 14 shows that the setup of. printers' type was. intended to indicate that Alaria (upper case i t a l i c s ) Del i s e i (spaced lower case roman) was to be transferred to Qjs_s§_ Deljgei (single spaced roman).  Thus, the nomenclatural type is the same as that of No. 6  above. 9.  Alaria dolichoraphis Kjellman, 1883a, p. 217, Pis. XXI, XXII, XXV. The type locality was stated to be Koljushin (eastern Siberia), and some  points just east of i t .  There is one sheet at Lund (LD), three at Stockholm  (S), and two at Uppsala (UPSV) labelled "Alaria dolichorachis Kjellm. F . R. Kjellman" in Kjellman's hand.  The date and locality of the Lund sheet is  "Sibirska Ishavet, PItlekaj 7.7.1879", of the Stockholm sheets "Sibirska Ishavet: Tschuktschlandets nordkust 18 7/7 79'% and of the Uppsala sheets "Arctiska Sibirien: Konyambay l a t . N„ 64° 4-9' Long. V. 172° 53'  18 §8-=30 79.»  Of these various specimens, only the one on the left side of the Lund sheet (LD 2251) is f e r t i l e , so i t is here designated as the lectotype. 10.  Alaria $&Xi£hQr3£&i£_ Kjellman f. longipes  Miyabe, I928, p. 957.  There was an implication by Nagai, .1940, p. I l l , that this was a recombination of Phasganpn longipes Ruprecht. I85I, p. 86 (No. 84 below). no  However,  reference was made to this in the original description, or in any subse-  quent publication of Miyabe himself, so i t seems best to consider i t a new epithet for typification purposes.  The type locality was state to be Nijo-  iwa and the east and west coasts of the Notoro Peninsula. At Hokkaido University (Fac. A g . , Sap.) there are four specimens identified as Alaria dolichorachis Kjellm. f. longipes. collected Aug. 9, I926, by J . Tokida "on  the sea shore" (drift?) at Hishitoma, Province Honto, Saghalien.  They are  also labelled "Stipe 29 cm. long", "stipe circa 45 cm.", "stip Qsic) 26 cm." and "stipe 29 cm. long".  At Berkeley (UC), there are two specimens labelled  "Alaria dollchorachis Kielim., f. longipes Miyabe Sept.- 1927. Kita-Shiranushi, Prov. Honto, Saghalien, J . Tokida f348)" (UC 54-3931,2). are  on the west shore of the Notoro Peninsula.  Both these localitie  The specimen annotated "stipe  circa 45 cm." (Fac. A g . , Sap.) is here designated the leetotype. 11..  Fucus esculentus  12.  Ceramium esculentum (L.) Stackhouse, 1797, P» 24.  13.  Musaefolia esculenta  14.  Orgyia esculenta  15.  Laminaria esculenta  16.  Phasgonon esculentum  17.  Agarutn esculentum  18.  PodopteriS [esculenta]  19.  Alaria esculenta  20.  Orgya esculenta Nos.  Linnaeus, 17^7? P» 135-  See page 51.  (L.) Stackhouse, 1809, p„ 53.  (L„) Stackhouse, l 8 l 6 , p. 8. (L.) C. Agardh, 1817, p. 13. (L.) Gray, 1821, p. 385.  (L.) Bory, 1826, p. 194. (L.)  De l a Pylaie, I829, p. 23.  (L.) Greville, I83O, p. 25, PI. IV, (L.) Trevisan, 1.845, p. 28,.  12 to 20 are a i l recombinations of £ . esculentus and so w i l l have  the same type. 21.  Alaria esculenta' f. fascleulata  Stroemfe.lt, 1886, p. 38.  The type locality was stated to be Eyrarbakki ( d r i f t ) .  There are three  specimens at Stockholm identified as this taxon with StroemfelVs signature, collected in the drift at Eyrarbakki, September, 1883.  The one annotated  "Alaria esculenta, (L.) Grev. f. fasciculate nob.I Is.1. Eyrarbakki, uppkastad i fjaeren Sept 1883 H.F.G. Stroemfelt" is here designated as the leetotype. (One  of the others lacks the "nob."; the other has, in addition, "costa  platt").  The designated specimen is the original of a photograph deposited  -100at Copenhagen (c), presumably implying that i t is the type. 22.  Fucus esculentus v a r . / ? minog Turner, 1802, p. 104. The sole reference under this name in the description was to F_. teres  Good, et Woodw., so both should have the same type. 23.  Pbasaonon esculentum jQ. minus,  See No. 63 below.  (Turner) Gray, 1821, p. 385, orth. mut.  The description here was merely an English translation of Turner's (No. 22 above), with the same sole reference to F_. teres, so the type should be the same also (see No. 63 below). 2k.  Pleuropterum fasciculatum  Yamada, 1935, P« 16, PI. 5.  The type locality was stated to be Yema, Urup I , at Hokkaido University (SAP 15356).  The type specimen is  It was collected and designated by  Yamada, and is identical with the illustration. 25.  Pleuropterum paradlseum f. fasciculatum  (Yamada) Miyabe et ISagai,  1933. P. 100. This was specifically a recombination of No. 2k above, and so has the same type. 26.  Alaria fistulosa  Postels £_. Ruprecht, lb%Q, p. 11, PI. 1.6.  27.  Orgva fistulosa  28.  Phasganon fistulosum (P. et R.) Ruprecht, 1851, p. 355.  See page 55.  (P. e_R.) Trevisan, 1845, p. 28.  Nos. 27 and 28 are recombinations of Alaria fistulopa and so w i l l have the same type. 29.  Alaria fistulosa f. platyphylla Setchell, i n Collins, Holden, et  Setchell, 1901, XLII. The type locality was stated to be the east shore of Amaknak I . , Unalaska Bay (in the Aleutian Islands).  These exsiccata were widely distribu-  ted, but since they consist of fragments of lamina and sporophylls, they could be portions of a very few individual plants.  A specimen at Berkeley (UC 9659^)  is here designated the leetotype.  It carries an annotation of Setchell's to  the effect that i t was sent to Kjellman for comment. 30.  Alaria fragilis  Saunders, 1901b, p. 425, P L L I V .  There is a specimen in the Farlow Herbarium annotated "A. fragilis N. sp?" in Saunders' hand, and also bearing an annotation by Kjellman referred to in Saunders' description.  This is here designated the leetotype.  cality of the leetotype is difficult to determine.  The l o -  Saunders' description give  three localities: Glacier Bay (80^); Prince William Sound (257); and Kukak Bay (333 1/3).  Kjellman's annotation on the specimen under discussion refers  to these three collection numbers, so presumably this plant is from one of these l o c a l i t i e s .  No. 31 below is described next by Saunders as "With the  last, Glacier Bay (79)"•  It is reasonable to presume that the type specimen  of Alaria fragilis f. fragilis is Saunders' " ( 8 o f ) " , from Glacier Bay. 31.  Alaria fragilis f. bullata  Saunders, 1901b, p. 425, PI. LIV.  The type locality was stated to be Glacier Bay (79).  Since i t is stated  by Saunders that several specimens were collected, a leetotype is designated here.  It is labelled in Saunders' hand "Alaria fragilis "bullosa Saund. No  79." and is also annotated In Kjellman's hand "No. 79 Alaria spec, nova bullata . . . . " 32.  Alaria grandifolia  J . G. Agardh, 1872, p. 26. See page 57-  33.  41arla pylajj X grandifolia  ( J . Ag.) Jonssen, 1904, p. 20.  This was an explicit recombination of AXarla grandjfplia and so has the same type. 34.  Alaria lanceo^ata  Kjellman, 1889, p. 39, PI. ?, figs. 1 - 3.  The type locality is stated to be Bering Island.  There are four sheets  at Uppsala (UPSV), one sheet at Lund (ID), and one sheet at Stockholm (S), labelled in Kjellman's hand "Alaria lanceolata Kjellm. mscr. 11-19 8 1879  Beringshaf. BeringBn F . E . Kjellman™..  A specimen at Lund (LD 2083) ap-  pears to be the original of the illustration, although a mirror image, and is here designated as the lectotype. 35.  Kjellman, 1889, p. 4 0 , PI. VI.  Alaria laticosta  The type locality was stated to be Bering Island (drift).  There is one  specimen labelled ".Alaria laticosta Kjellm. Beringshaf. Beringon 18  79 o  F. R. Kjellman" in Kjellman's hand In the herbarium at Uppsala (UPSV).  Since  i t was stated that only one specimen was collected, this specimen must be the holotype. 36.  Stroemfelt, 1886, p. 3 8 .  Alaria linearis  The type locality was stated to be the mouth of ReySarfjox&ur, off .Karlskali farm and at Seley.  There are five sheets at Stockholm (S), one  from Karlskali and four from Seley, collected 29/30 Jane, 1883, by Stroemfelt.  The plant from Karlskali, since i t is labelled "Alaria, linearis nob."  (as opposed to "Alaria linearis Stromfelt"), is designated here as the lectotype. 37.  Phasganon macropterum Ruprecht, 1851, p. 353. The localities given by Ruprecht are Msmgabal, Asa Island, and damaged  specimens from Cape Mchta, Great Schantar I . and Dshukdshandran„ of specimens were distributed by Ruprecht from St. Petersburg. these is now at each of  A number  One each of  Farlow (FH), Kew.(K), Paris (PC), Lund (LD),.  Stockholm (S), and Dublin (TGD). Acad. Petrop. Mare Oohotzk".  They a l l have a printed label reading "Herb.  Those at Dublin, Lund, Paris, and Kew have  printed in addition "Exped. S i b i r . Acad. 1844" and "FJaas.ganon_ macropterum R" in Ruprecht's hand. Alaria macroptera".  Those at Farlow and Stockholm have "leg. Middendorf  It is best to choose a lectotype from the former group,  as these have a definite date previous to Ruprecht s description. 1  The  -103-  specimen at Lund (LD' 2085) is here designated the lectotype, 38.  Alaria macroptera  (Ruprecht) Yendo, 1919, p. 79.  This was an explicit recombination of No. 37 above, and so has the same type. 39=  Alaria. marginata Postels et Ruprecht, 1840, p. 11 „ See page 59.  40.  Orgya marginata  41.  Phasganon marginatum (P. g t R . ) Ruprecht, 1851, p. 355.  (P„ gt R.) Trevisan, 1845, p. 28.  Hos. 40 and 41 are explicit recombinations of Alaria marginata and so w i l l have the same type. 42.  Laminaria musaefolia  De l a Pylaie, 1829, p. 31.  The type locality was stated to be St. Pierre.  There is one sheet at  Paris (PC) which is annotated: "Laminaria Mqsaefolia Dip. f l . T . I . p. 31: Fuc. esculentus Linn. var. ignota et in stipem geminam constituenda Fuc. esculentus . Lightf. F l . Scot. Podopteris musaefolia Dip. Mss. r e c e u i l l i en l8l6 et 1820: c'est 1'espece l a plus abondante dans l a rade de l'^le St. Pierre." in De l a Pylaie's hand. 43.  Alaria musaefolia  It is here designated as the lectotype. (De l a Pylaie) J . Agardh, 1872, p. 23.  This was an explicit recombination of  musaefolia and so i t  has the same type. 44.  Alaria esculenta f. musaefolia  (De l a Pyl.) Kjellman, 1883a, p. 212.  This was an explicit recombination of Alaria musaefplja, so i t has the same type. 45.  Laminaris-  musaefolia  remotifolia  De l a Pylaie, I829, p. 35.  A type locality was not given for this subspecies, so presumably It is St. Pierre for this variety as well as for F_„ musaefolia. ,An incomplete plant at Paris (PC), with the missing parts-sketched In, agrees well with the description, and is annotated: "Laminaria musaefolia var. I'emotifolia D.L..P. F l . Terr. Nov. p. 35. Pinnules appartenant a cette espece que j ' a i rencontrees dans ma collection parmi divers fucus flue je n'ai prepares quand . . . . . . . . [illegible] mon retour en France l i e St. Pierre l8l6."  -104probably in De l a Pylaie's hand. 46. Alaria nana 47.  It is here designated as the leetotype.  Schrader, 1903, p. 157, Pis. 24 - 2 6 .  Alaria praelonga f. nana  See page 62.  (Schrader) Setchell, in Collins, Holden,  et Setchell, 1905, Ho. 1292. 48. Alaria marginata f. nana  (Schrader) Collins, 1913, P° 110.  Nos. 47 and 48 are recombinations of Alaria nana and so w i l l have the same type. 49.  Alaria oblonga  Kjellman, 1883a, p. 220, Pis. XXII and XXV (figs. 21-24).  The type locality was stated to be Irkaypi.  Kjellman's translitera-  tions of 19th century Russian place names into old Swedish are sometimes difficult to identify. Siberia.  This one appears to be Ryrkurpiya, in northeast  There are five sheets at Stockholm (S), and one sheet at Lund (LD  2084), labelled "Alaria oblonga Kjellm. Sibiriska Ishavet. dates] 1878 F.R. Kjellman" in Kjellman's hand.  Irkajpi [various  The specimen at Lund (LD  2084) Is here designated the leetotype. 50.  Alaria ochotensis  Yendo, 1919, P» 84, Pis. I l l , XIX. See page 64.  51. . Pleuropterum -paradiseum Miyabe et Nagai, 1932, p. 127, figs. 1, 2. See page 65. 52.  Pleuropterum paradiseum f. brevipes  Miyabe et, Nagai, 1933, P. 101.  The type locality was stated to be Broughton Bay, Shumushir I. (in the middle Kurile Islands).  There are two plants at Hokkaido University  (Fac. Ag ., Sap.) labelled "Pleuropterum paradiseum Miyabe et Nagai f. bre;  vipeg Broughton Bay Shumushir July 22, 1930 M.N." in Nagai's hand. missing the holdfast (which is an important defect when the character is the shortness of the stipe) so the other here designated as the leetotype. 53-  Orgyia pinnata  (Gunnerus) Ghobi, 1878a, p. 77.  One is  subspecific  complete specimen is  -ic-5This combination does not seem to nave been published by Rostafinski, as i t is sometimes cited in herbarium annotations, although a reference, is made to a letter of Rostafinski in Ghobi's l i s t of synonyms. that Rostafinski's manuscript postdates his own.  Ghobi states  Previa pinnata is a recom-  bination of Fueus pinnatus Gunn. (No. 103 below) which, while an illegitimate name, may have priority of Fucus esculentus L. (see p. 113 below).  These two  species are obviously synonymous; this is the f i r s t combination of the epithet ninnatus with the name of one of the genera synonymous to Alaria. Therefore, the well established Alaria esculenta (L.) Grev. has the priority. The typification of No. 53 should be the same as that of No. 103 below. There is one sheet in Gunnerus' herbarium at Trondheim, which is labelled "Gunnerus' hb. A. (forsynet med nr. svarende t i l F l . Norv. No. 313 Fupus esculentus Linn)".  It would be d i f f i c u l t to prove that this is actually  type material for the Flora Norvegica, but this is not important, since Fucus pinnatus Gunn. Is illegitimate.  This plant is designated the neotype  of No. 53; i t might have been consulted by Ghobi or Rostafinski.  It can be  designated definitely as the leetotype of No.. 54, as i t agrees with Foslies' description of a plant in Gunnerus' herbarium. 54.  Alaria esculenta f. pinnata (Gunnerus) Foslie, 1886, p. 114. See note con T No. i53 above. r  55.  Alaria praelonga  56.  Agarum p y l a i i  57.  Laminaria p y l a i i  58.  Alaria p y l a i i  59-  Orgya p y l a i i  Kjellman, 1889, p. 38, PI. I V .  Bory, 1826, p. 194.  See page 66.  See page 6:8.  (Bory) De l a Pylaie, 1829, p. 31°  (Bory) Greville, 1830, p. xxxix. (Bory) Trevisan, 1845, p. 28.  Nos. 57 "bo 59 are explicit recombinations of Agarum p y l a i i . and so w i l l have the same type.  -10660.  Alaria taeniata  Kjellman, 1 8 8 9 , p. 3 6 , PI. 7, figs. 1 - 3 .  61.  Alaria tenuifolia  See page 7 0 .  Setchell, in, Collins, Holden, et, Setchell, 1 9 0 1 , XLV.  See page 7 1 , 62.  Setchell §t Gardner, 1 9 0 3 , p. 2 7 4 .  Alaria tenuifolia f. amplior  Localities and collections cited in the description are "Esquirnalt, B. C. J_.2_.fiL., .No.  645'.,  _L..4._l. No. 1 8 7 5 ' . ; Hear Roach [sic] Harbor, San Juan  Island, Wash., _l.I_.__,., Nos.  6 5 O ' , 65I ." 1  A specimen at Berkeley (UC 9 6 6 5 1 ) ,  labelled "Algae of Puget. Sound No. 64.5 Alai^ta. tenuifolia f „ amplior Esquimalt, B. C. N. L . Gardner May 5 , 1 9 0 1 " and annotated "Sent to Kjellman and returned cf. l i t July 3 0 , 1 9 0 2 " in Setchell. s hand, is here designated as the lectotype. 1  63.  Fucus teres  Goodenough gj_. Woodward, 1 7 9 7 , . P« 2 7 4 .  The type locality was stated to be Holyhead, and the type material was collected by "the Rev. Mr. Davies".  There is one sheet at Kew (K) stamped  "Ex. Herb. Dr. Goodenough" and labelled "Fucus teres Isle of Anglesea Reed. Hugh Davies" in what may be Goodenough•s hand.  While the remounting of the  specimens at Kew has introduced an element of uncertainty into the identity of labels and plants, in cases such as this where there are more than one plant and label on a sheet, these plants would appear to be good authentic material.  The plant mounted on the right hand side of the sheet is here desig-  nated as the lectotype. 64.  Alaria valida  Kjellman et Setchell, in Setchell and Gardner, 1 9 0 3 , p.  2 7 8 , PI. 2 1 .  The localities cited are Unga, Alaska, and Whidbey Island.  Two specimens  are designated as co-types at Berkeley, one (UC 9 6 6 6 3 ) from Whidbey Island, and the other (UC 9 6 6 7 3 ) from Unga.  They are very different in form.  The  specimen from Whidbey Island (UC 96663) is here designated as the lectotype because Plate 2 1 was drawn from i t .  -10765.  Alaria valida f. longipes  Setchell,, in Setchell et Gardner, 1903, p. 278.  The type locality was stated to he the west shore of Whidbey Island, and the collection number l.L.G;. 111.-  This description applies to three sheets  at Berkeley (UC 132634, 96667, and 96666)0  One of these (UC 96666) best  corresponds to the description, so i t is here designated as the leetotype. 66.  Alaria yezoensls  Miyabe, 1902, p. 54, PI. XXIII.  (Alaria yessoensis  according to Yendo, 1919, p. 86.) The type locality was stated to be Hidaka and Nemuro Provinces, Hokkaido (Japan).  Miyabe s illustrations were done by a professional artist and seem 1  to be somewhat idealized.  A plant at Hokkaido University (Fac. A g . , Sap.),  labelled: "Maria yezoensls Miyabe K2 No. 216 The cross-section of the midrib same as KI. Kombu Mori nr. Kushiro Miyabe VII 24, 1894" in Miyabe s hand, and annotated "Alaria praelonga Kjellm, Det Yendo" in 1  Yendo's hand seems to be the most suitable specimen available.  It was i n -  volved both in Miyabe's original description and Yendo's reduction of the species to synonymy under 4. praelonga.  This specimen is here designated, the  leetotype. Names for which type material found was inadequate; 67.  Alaria cordata  Tilden, American Algae No. 24l, 1898.  The material distributed by Tilden is heterogeneous, consisting of the laminae of immature Alaria tenuifolia from Minnesota Reef (San Juan Islands) and sporophylls of Alaria marginata from Port Renfrew, B. C. A lamina ident i f i e d in Tilden's hand is selected as the leetotype (MIN 16748). 68.  Alaria e l l i p t i c a  Kjellman, 1883a, p. 221, PI. XXIII.  Localities were stated to be Koljushin and Pitleikaj (northeast Siberia). There is one sheet at Lund (LD) labelled "Alaria eAliptica_ Kjellm. Sibiriska  .-108Ishafvet, Pitlekaj juni 1879 F„ R. Kjellman" and one at- Stockholm (S), l a helled in Kjellman's hand "Alaria e l l i p t i c a Kjellm. Sibiriska Ishavet: Koljutschinfjorden 18 2l/6 79 F . R. Kjellman".  Both specimens are i n f e r t i l e .  The Lund specimen (LD 2082) has been designated the ^Leeto] type by Yendo. 69.  Kjellman, 1883a, p. 212.  Alaria esculenta f. australls  The fact that Kjellman referred to Alaria esculenta "f, typica" i n Plate 25 of the same publication, indicates that i t was not Kjellman's intention to include the type in either f  australls or f., musaefolia.  It is ambiguous  on p. 212 as to whether this is the case, or whether he intended to divide Alaria esculenta in its entirety between f. australls and f. musaefolia. Alaria esculenta f. australls cannot be typified on any of Kjellman s speci11  mens, since the description is only a citation of J . Agardh's description of Alaria esculenta i t s e l f (1872, p. 22).  So f. australls is typified on the  cited illustration, Turner, 1802, PI. 117, which is the leetotype. 70.  Laminaria esculen£p, var. n o l t i i  Hornemann, I.837, p. 737.  Hornemann appears to have made the f i r s t valid publication of this taxon in I837, since in 1827, p. 97,  Q  e  proposed no epithet with the description.  It was noted that this plant was collected by Professor Nolte at F^hr and Gult (North Friesian Islands).  There is a juvenile specimen at Copenhagen  (C) labelled "Laminaria . v u m Doctor Nolte... {illegiblej'" in a hand that is probably Hornemann's, and a later annotation "Fbhr 1820 P. Anderson". This specimen is here designated the leetotype. 71.  Alaria esculenta y-pinnatlfida Postels ei. Ruprecht, 1840, p. 11. There is one sheet at Farlow (FH), labelled "Herb. Acad. Petrop.  Alaria esculenta v. pjnnatifjLda PR Ad Portam St. Petri et Pauli Camtschatkae Exped. Iiitk." in Ruprecht's hand.  There is no indication that there is a  better specimen at Leningrad (A. D. Zinova, pers. comm.), so the Farlow  -109specimen is designated the lectotype. 72.  Alaria esculenta ^ l a t i f o l i a  It is a lamina fragment.  Postels et Ruprecht, 1840, p„ 11, PI. XVII.  The type locality is not definitely stated, hut It is probably In the North Pacific at Sitka, Unalaska, or Petropavlovsk (Kamtschatka)„  No speci-  men could-be found, which could be considered authentic material, in the herbaria visited, but there is one specimen in poor shape at Leningrad (IE) labelled. "Alaria esculenta var. l a t i f o l j a forma singularip Spec. paras i t i c a . Spiraliter contorta insidente Lam, digitata. Kamtschatka ad portum St. Petri et Pau.ll, Expeditio liltk." according to A. D. Zinova (pers. comm.).  The "forma singularis" probably  refers to a solitary specimen, but since Yendo (1919, P° 83) refers to se-  s  veral specimens, this particular one is designated the lectotype. 73.  Alaria flagellaris Stroemfelt. 1886, p. 4 l , PI. I I . The localities are stated to be Seley and Karlskali, Iceland.  There  are two sheets at Stockholm (S) labelled "Alaria flageLlaris nob." in Stroemfelt's hand, one from Seley and one from Karlskali, collected on the 29 and 30 June, I883,. respectively. shape.  They are both immature and generally in poor  There is a photograph of the Seley plant deposited at Copenhagen (c),  presumably implying that i t is the type.  The plant from Seley is here desig-  nated the lectotype. 74. Alaria macrophylla Miyabe. 1902, p. 56, PI. 25. The type locality was stated to be Kunashiri I. (South Kurile Islands). No specimen from Kunashiri I . , collected before 1902, could be found at Hokkaido University, so the illustration is here designated as the lectotype. 75.  Alaria^ membranacea J . Agardh, 1872, p. 26. The type locality was stated to be Spitsbergen, and the -type collection  made by Berggren.  There is one sheet at land, labelled "Spitsbergen, Berggren  -noAlaria membranacea J . Agardh Ag. Gr." somewhat i l l e g i b l y in Agardh's hand. A l l the plants are immature. The plant mounted on the right hand side of the sheet (ID 2113) is,the best developed and is here designated the lectotype. 76.  Alaria pylaij, ftmembranacea ( j . Ag.) Rosenvinge, 1893, p. 838. This was an explicit recombination of Alaria membranacea. and so has the  same type. 77.  Alaria platyrhiza Kjellman, 1906, p. 11, PI. II, f i g s „ 11-13. The type locality was stated to be Jan Mayen Island, with only one spe-  cimen seen.  There is one plant at Uppsala, ,(UPSV) labelled:  "Jan Mayen 8/8 00 30 m. Alaria, platyrhiza Kiellm. (In Kenntn. d. mar. a l g f l . Jan Mayen 1906) E . H. Gran." which corresponds exactly to Kjellman's i l l u s t r a t i o n . cimen.  It is a juvenile spe-  Since Kjellman states that his description is based on only one spe-  cimen, this must be the holotype. Names for which no t^Qoe, material was found. 78.  Alaria esculenta c^angustifolia  Postels et Ruprecht, 1840, p. 11.  No type locality was given for this variety. 79.  Laminaria esculenta var. platyphylla De l a Pylaie, 1824, p. 178. The locality.was stated to be St. Pierre.  80.  TapHnarla.  esculenta var. polyphylla De l a Pylaie, 1824, p. 179.  The locality was stated to be the extremity of lower Brittany. 81.  Laminaria esculenta var. remotifolia  De l a Pylaie, 1824, p. I79.  No locality was given for this variety.  The type specimen is probably the  same as No. 45 above, but i t appears impossible to prove this. 82.  Laminaria esculenta var. taeniata  De l a Pylaie, 1824, p. 179.  No locality was given for this variety.  -111-  83.  Fucus fimbriatus  Gmelin,  I768,  p„ 2 0 0 .  The locality was stated to he Kamschatka. 84.  Phasganon longipes  Ruprecht, I 8 5 I , p. 3 5 3 .  The locality was stated to be the Qchotsk Sea. 85.  Alarja marginata  musaeformis  Postels et Ruprecht, 1 8 4 0 , p. 1 1 .  No locality was given.for this variety. 86.  Laminaria musaefolia var. decipiens  De l a Pylaie, 1829, p. 3 6 .  The locality was stated to be Newfoundland. 87.  Fucu^ polyphvllus  Gmelin, 1 7 6 8 , p. 2 0 6 .  The locality was stated to be Kamschatka. 88.  Phasganon scotieum  Walker, in N e i l l , 1 8 1 5 , p. 4 „  No locality was given for this species. Names which are illegitimate or removed from the genus. 89.  Phasganon alatum Ruprecht, 185.I, p. 3 5 3 . Ruprecht suggested this epithet as a preferable synonym for Fucus  esculentus L . , using considerations of pre-Linnaean priority and the etymological suitability of epithets, not recognized by the Code. is superfluous. 90.  (Article 2.)-  Alaria amplexicaulls Yendo  (1919,  So, Ph. alatum  p.  135)  Martens, 1 8 8 6 , p. 144. removed this species to Undaria.  Reference to the  original description clearly shows that this decision was correct. 91.  Maria, caulescens Savage,  1945,  p. 200.  This, Savage stated to be an annotation on the sole specimen of Alaria in the Linnaean Herbarium.  Examination of the specimen indicated that this  must have been a misreading from a photograph of an annotation "Alaria esculenta." 92.  Alaria costata  Fallis, 1916, PI. 25, fig. 1.  This is apparently a typographic error by F a l l i s , referring to Alaria cordata Tilden. 93.  Laminaria despreauxii  Bory, in Ruprecht, 1 8 5 1 , p. 3 5 7 .  This is a reference by Ruprecht to a herbarium specimen.  Although the  writer found a number of similar herbarium annotations, this seems to have been the f i r s t time this epithet appeared in print, even as a nomen nudum. 94.  Alarja despreauxii  (Bory, i n Ruprecht) J . Agardh, in Nordenskiold, 1 8 7 2 ,  p. 1 0 8 1 . This appears to be the f i r s t time that this epithet was combined with the genus Alaria. and is a nomen nudum. 95.  ALarLa. dplichoracMs f. typica  Nagai, 1940, p. 1 1 0 .  This is to be rejected under Article 24 of the Code. 96.  Alaria esculenta f. typica  Kjellman, 1 8 8 3 , p. 341.  This is to be rejected under Article 2k of the Code. 97.  P.  Alaria_ esculenta var. l a t i f o l l a f. singularis  Ruprecht, in Yendo, I9.I9,  89.  See the note under No. 7 2 above. 98.  Alaria fistulosa f. stenophylla  Setchell, in, Collins, Holden, et  Setchell, 1 9 0 7 , B, XLIII. Alaria fistulosa was divided by Setchell into f. stenonhylla and f, platyghylla without reference to the type, mostly on a basis of a narrow and a wide midrib.  Comparison of material, distributed by Setchell as belonging  to these two forms, with the type specimen of A. fistulosa f. fistulosa i n d i cates that f. stenophylla includes the type and so should be rejected under Article 2 6 of the Code. 99.  Laminaria linearjl.s  De l a Pylaie, .1829, p. 3 7 .  This Is clearly an Alaria according to the original description, but i t  -113does not seem to have been formally combined under this genus before Stroemf e l t ' s description of Alaria linearis (No. 36 above) and so lacks priority to Thus, No. 99 becomes a later homonym.  i t in the genus A l a r i a . 100.  laminaria meaalopteris  De l a Pylaie, 1829, p. 29.  This is cited as a synonym of laminaria p.vlaii and is to be rejected under Article 3^ (l) of the Code.  De l a Pylaie's motive in l i s t i n g i t seems  to have been to indicate that his manuscript description of the new taxon was prior to Bory's publication of i t . 101.  laminaria n o l t i i  (Hornemann) J . Agardh,  Yendo, 1919, P- 109.  This was merely mentioned by Yendo as a herbarium annotation. a recombination of No. 70 above.  It is  It is to be rejected under Article 3^ (3)  of the Code.  102.  Alaria pinnatifida  Harvey,  This was removed by Yendo  i860, p. 329.  (1919, p. 135) to Undarla. Reference to the  original description and material clearly shows this is justified. 103.  Fucus pinnatus  Gunnerus, 1?66, p. 96.  This is illegitimate as a later homonym of Fucus pinnatus Hudson, I762, p.' k 6 k , which is a red alga.  While Gunnerus' species apparently has  the priority over Fucus esculentus L,, its obvious synonym, there is some evidence that parts of the Flora Norvegica were published later than the date in the t i t l e page (Dixon,  i960, p. 303).  priority over Fucup P i n n a t u s Gunnerus.  Then Fucus esculentus L. would have However, this is an academic ques-  tion here, since F_. pinnatus Gunn. is illegitimate in any case. 10k.  Alaria pylaij f. typica  Rosenvinge, 1893, p. 838.  This is to be rejected under Article 2 k of the Code. 105.  Alaria striata  J . Agardh, in Yendo, 1919, p. 93.  This is mentioned by Yendo as a herbarium annotation.  -114"  106.  Alaria, tenuifolia f. typica  Setchell, in. Setchell gt Gardner, 1903,  P. 272. This is to he rejected under Article 2k of the Code. 107.  Fucus tetragonus  Goodenough et Woodward, 1797, p. l40.  Goodenough and Woodward divided Fucus esculentus L . Into .£,. teres and F_„ tetragonus without reference to the type. The l o c a l i t y of F . teres is Holyhead, and that,of F . tetragonus is the Isle of Man.  Goodenough and Woodward  stated that they were endebted to "the Rev. Mr. Davies" for apprising them of the differences between the two. E» flmbrlatus Gmelin.  The only reference under F_. tetragonus is  The absence of any authentic material of F . tetragonus.  and the fact that the Rev. Davies lived near Holyhead, not the Isle of Man, indicates that Goodenough and Woodward probably never examined material of E . tetragonus;  It is so vague an entity that i t cannot be considered as  excluding the type, and is to be rejected under Article 2.6 of the.Gode.  -1152 - Synopsis of taxonomic c r i t e r i a .  \  Entity*  Criteria of original description.  1.  Compressed rachis; rectangular midrib; short, terete stipe.  2.  Lamina with rounded base, fine transvers corrugations.  3.  Thick sporophylls; flattened rachis; e l l i p t i c a l midrib.  4.  Long, narrow, numerous, fasciculate sporophylls.  5.  Broad midrib; short and robust stipe.  6.  Widely separated, laurel-leaf-like sporophylls.  7. & 8 .  Recombinations of 6 .  9«  Broad, prominent midribj long, flattened rachis.  10.  Long stipe, 30 to 50 cm. long.  11.  Type species.  '12. - 20.  Recombinations of 11.  21.  Long, narrow, fasciculate sporophylls.  22.  Lamina with attenuated base. .?  23.  Recombination and orthographic variant of 22.  24.  Wide midrib; fasciculate sporophylls, some typically sterile.  25.  Recombination of 24.  26.  Midrib including air chambers.  27. 8s 2 8 .  Recombinations of 26.  29.  Wide lamina and midrib} numerous narrow sporophylls.  30.  Long stipe; sporophylls few and widely separated.  31.  Bullate lamina.  32.  Long stipe; narrow and linear sporophylls.  33.  Recombination of 32.  34.  Broad lamina; linear sporophylls; inconspicuous midrib.  35«  Very broad midrib; short thin sporophylls. * See page 19.  1 -  J  -116Entity  Criteria of original description  36.•  Narrow lamina; rachis undifferentiated from stipe.  37•  Broad lamina base; large crowded sporophylls, membranaceous distally.  38.  Recombination of 37.  39«  Sporophylls linear, rounded at apex and base, petiolaate, with a clear smooth margin.  40. & 41.  Recombinations of 39.  42.  Oval, lanceolate lamina; long stipe; falcate sporophylls.  43. 8s 44.  Recombinations of 42.  45.  Sporophylls distant, lanceolate, with cuneate bases.  46.  Stipe robust, terete; rachis tapering; long, narrow, sporophylls.  47. 86 48.  Recombinations of 46.  49.  Elongate, linear-oblong lamina.  50.  Persistent sporophyll petioles; branched gland cells in blade cortex.  51.  Some sporophylls sterile, with a midrib.  52.  Short stipe.  53'  Recombination of Fucus pinnatus Gunn., synonym of No. 11.  54.  Recombination of 53'  55 • 56.  Attenuated lamina; sporophylls variable in form. Ovate lamina; sporophylls cuneate, undulate, much enlarged distallly.  57• - 59•  Recombinations of above.  60.  Narrow lamina; midrib rectangular, contracted at blade.  61.  Stipe long, flattened; blade membranaceous.  62.  Sporophylls wide, with sori restricted to proximal third.  63.  Terete midrib; membranaceous sporophylls; lamina attenuate.  64.  Stipe short; sporophylls long, linear-lanceolatej midrib wide.  -117Criteria of original description Stipe long. Rachis short, flattened; sporophylls long-linear, rounded at apex.  r  -118Appendix II - Statistical data Table 1. - Descriptive data of samples used. CO CQ fD P  o  53 O >i B O H (B ,_. • H 0> I— CD 1  o  ra  S3  O H• O  O 0  6.  4029  27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 4o.  5279 5500 5501 5190 5503 5504 5505 5180 5183  7. 4069 8. 4035 9. 4o65 10. 4092 11. 4238 12. 440813. 4242 14. 4410 15. 4076 16. 4244 17. 4031 18. 4589 19. 5001 20. 5065 21. 5155 22. 5201/2 23. 5248 24. 5273 25. 5275 26. 5277  42.  B fD  M  N CD H CD  H-  1. 3856 2. 2682 3- 3271 4. 3274 5. 4051  41.  a) c  CQ CO !»  Remarks 1,3. Duplicate. 1,3Duplicate. 1,3Duplicate 1,31,3. 1,3. Duplicate. 1,3. 1,31,2,3. Duplicate. Duplicate. 1,3Duplicate. 1,3. 1,3. Duplicate. 1,3. 1,2,31,31,31-31,360 Western C h a n n e l 1,2,360 Nr. Sheet Hbr. 1,3 Duplicate. 60 Wee cumbrae I. 1,2,3. Duplicate. 6 l Wee Cumbrae I . 1,2,3. 6 l Muroran Duplicate. 1,3. 6 l Ainukappu Rocks 1,2,3. 6 l Hakodate 1,3. 6 l Cape Erimo 1,2,3. Duplicate, 62 Ainukappu Rocks Courtesy Y Yamada. VII 62 Daikokujima 1,3. c/o Y. Yamada XI 62 Brockton Pt. 1,3. Duplicate. XI 62 Brockton Pt. 1,3« Duplicate. X 62 Pt. Lepreau •1,3. c/o R. Taylor VIII 62 Nr. Newport, Ore. 1,3VII 62 Friday Harbor 1,3. VI 62 Port Renfrew 1,3V 6 l Farland Point 1,3- Duplicate VII 6 l Farland Pt. 1,3- Duplicate. 3 added to No. 4 — 2,311 added to No. 13 — 2,3. 28 added to No. 29 — 2,35 enlarged - 2,38 enlarged — 2,315 enlarged 2,3. 17 enlarged — 2,3.  18 17 13 12 15 13 10 21 25 18 13 20 14 25 17 11 18 18 25 15 25 25 15 25 15 25  marginata nana marginata marginata esculenta crispa crispa tenuifolia crispa crispa taeniata marginata taeniata nana praelonga fistulosa fistulosa nana esculenta esculenta esculenta crassifolia angusta crassifolia crassifolia angusta  II 29 V 18 VI 18 VI 10 VI 8 VI 12 VI 9 VI 12 VI 14 VI 27 VI 7 VII 27 VI 7 VII 12 VI 27 VI 8 VI 13 VII 22 IX 24 X 22 III 23 IX 25 X 25 XI 9 XI 2 VII  15 15 15 15 15 15 15 12 12 25 25 25 21 24 21 25  an; tenuifolia tenuifolia esculenta marginata tenuifolia marginata esculenta esculenta — — -— — — --  17 3 5 19 No. No. No. No. No. No. No.  60 Sandstone Creek 59 Kains I . 59 Eagle Pt. 59 Eagle Pt. 60 Casco Bay, Attu 60 Murder P t . , Attu 60 Cape Agagdak 60 Chichagof Bay 60 Cape Agagdak 60 Trappers Cove 60 Seldovia Pt. 60 Ocean Cape. 60 Seldovia Pt. 60 Ocean Cape 60 Cape Agagdak 60 Seldovia Pt. 60 Murder P t . , Attu  -119Table 1. (Cont.) CO CQ 53 O O CD  P  H ft 23 O  o  H H CD O  sic: C co  CQ 01 f» CD  P CD f-  H CD  O H-  • a  c+ H-  Species:A.  o  43. 44. 5052 45. 5028,31 46. 5116 47. 5150 48. 5132 49. 5l4o 50. 5148 51. 52. 5139/ 5141 53. 5120 54. 5143 55. 5145 56. 5147 57- 5191 58. 5192 59. 5193 60. 5155 61. 62. 63. 5057 64. 4590 65. 5506 66. 5507 67. 2798 68. 4203 69. 4202 70. 4073 71. 5248 72. 4054 73. 4191 74. 3276 75. 3274 76. 5262 77. 5279 78. 4267/ 4295 79. ^378/ 4387 80. 5081 81. 5144 82. 5001 83. 5057 ' 84. 5058  Locality  Date  25 20 20 25 25 20 25 21 22 24  — No. 34 added to No. 35— X 60 Reykjavik esculenta 6 X 60 Akranes esculenta 4 V 6l Coosvaddig esculenta VII 6l Eddystone Rk. esculenta VI 6l Kgistiansund esculenta VI 6l Maloy esculenta VI 6 l Kristiansund esculenta — No. 30 enlarged — VII 61 Maloy esculenta  21 20 20 21 25 25 22 25 20 24 25 6 !5 15 15 9 6 6 15 9 15 11 10 3 8 5  esculenta esculenta esculenta esculenta esculenta esculenta esculenta esculenta  V VII V V XI XI XI  6l 61 6l 6l 62 62  Cape Cornwall S c i l l y Is. Fife Ness Berwiek-on-Tweed Raunfjord Fanafjord v  62 KorsfJord  22 III 6l Wee Cumbrae I . No. 10 enlarged — No. 18 enlarged 19 XI 60 Cor Eileann esculenta 13 VII 60 Western Channel igjjgn^ VI 62 Albert Head tenuifolia; 5 4 II 62 Botany Beach nana VI 59 Botany Beach nana praelonga 23 VI 60 Cape Sarichef fistulosa 23 VI 60 Cape Sarichef fistulosa 12 VI 60 Cape Agagdak angusta 25 X 6 l Ainukappu Rocks 10 VI 60 Casco Bay, Attu. erispa 22 VI 60 Cape Sarichef erispa tenuifolia 18 VI 59 Turn Pt. tenuifolia 18 VI 59 Roche Hbr. 23 XI 6l Abashiri fistulosa 4 VII 62 Daikokujima praelonga 28 VI 60 English Bay taeniata  5 taeniata  25  VI  60 Pasagshak Pt.  13 esculenta 13 esculenta 9 esculenta  22  VI V IX  6l Mull of Kintyre 6l Helmsdale 6l Nr. Sheet Hbr.  6 esculenta 8 esculenta  19 l6  XI 60 Cor Eileann I -6l Wee Cumbrae I.  Remarks 2,32,3. 2,3. 2,32,3. c/o M. Parke 2,3. 2.3. 2,3. 2,3. 2,3. 2,3. 2,3. c/o M. Parke 2,3. 2,32,3. c/o H. Brattstrom. 2,3. 2,3. 2, Duplicate. 2,3. 2,3. 2,33. 33. 333. 3. 3. 3. 3. 3. 3. drift, 3. 3. c /o Y. Yamada Resoaked 3material Resoaked 3. material 3. 3. surplus not 3used before 33-  -120Table 1. (Gont.)  C OC QS3 oo C D l» H o H D V> H < a  t—"  3 O  85. 86. 87. 88. 89. 90. 91.  CQ C D  o 5155 U136 4o8l 1+587 3398  92. 3327 93. 3^97 9k. 2761 95. 3275 96. 3792 97. 3855 98. 3851 99. 3805 IOO.3810 101. 3847 102. 3833 103. 3901/ 3921 104. 105. 5155 106.  C cn O H>  Remarks 22 III 6 l Wee Cumbrae I. 3. dried 7 esculenta various dried plants from Aleutian Is. — 20 same plants as No. 86 resoaked --3« 20 15 VI 60 Zeto Pt. 3. I erispa 13 VI 60 Finger Bay, Adak. 3. 8 taeniata 12 VII 60 Halibut Point 3. 7 taeniata 22 VII 59 Loran Station, 3. II marginata Spring I. 21 VII 59 Big Bay, Spring I.3. 11 marginata 24 VII 59 MacLean I. 3. 10 marginata 4 VI 59 Kelsey Bay 3. 13 nana VII 59 Mt. Dallas Beach 3. 11 marginata II 60 Ogden Pt. '.; . 3* .Intermediate 10 tenuifolia Breakwater with marginata. II 60 Otter Point 3« 5 nana II 60 Otter Point 35 nana II 60 Beacon H i l l Beach 3. 11 tenuifolia II 60 Ogden Pt. Breaks?. 3. 7 tenuifolia II 60 Glacier Pt. 3« 21 nana II 60 Glacier Pt. 39 nana V 60 Brooks Peninsula 3. 25 marginata 6  — dried plants of species not included in the statistical analysis — 7 — measurements of No. 85 multiplied by a correction factor — 20 measurements of No. 87 multiplied by a correction factor - -  -121-  \  Table 2 - Data concerning duplicate samples. Serial No. of first sample  Serial No. of second sample  34  35  2.2  Same individuals, different time.  20 23  21 26  4.4 5.2  Same locality, level; different time.  3 28  4 29  2.3 1.0  Same time, locality, level; different samples  11 7  .13... 9  3.0 2.3  Same time, locality; different level.  1 9  33 10  4.1 4.7  Same time, area, level; different exposure " to surf.  21  6o  0.4  Same sample; duplicate measurements.  Nature of duplication  n  ti  ti  ti  11  II  ti  VI  ti  11  VI  H  II  II  II  tl  - 122 H CM fO-dco  CO ONCO t - 0 0  -* co  vo  C O  VO  co OJ co H  q t-vo t-  rH  ONLTNVO-d-  o H O  rH  L T N C O C O C—  ON  t— r-j VO VO r-j g  cOJ VO OJ  LC\  CM -3OJ  ro OJ  OJ O OJ ON  H  <f  rH  ^ ^ H r ^ H  g  rH rH  g  g  vo vo  H  LTN  C— ON CO O C\j  LTN C O L T N C O L O J V Q C O C O V O -4"  LTN  CO  ITN  rj  oi  H  ajONHHcH  vo 0 0 0 0  rH  Q  H  Q H  OJ  C- O N C O  CO  ONCO  vo C O C O C O ^ t  t^-d- O rH  LO. O N C O C O L O J C O L O . H  vo  UP  [--vo  Q  Q  CO  OJ  t> vo Lf\ t-vo CO  ON  r-j g C O ON rH rH rH  0 \ H 0 4 ON  q  ON  cy o q H  H i—i H  U-NCO  LT\VO ON ON ONVO  rH rH  3^33 3  O H O \ O H H  t-- oj ON  C— ir\ C  C O H  H  o  '  H  L T N - = f H H  H H n rH rH rH rH g H vo co C O V O I A U N C O ] rH r—I  t-vo c-  LT\  ON ON^t  H r-i o  CO CO  rH rH  0 0 Vvo t > - L O , rH VO O LTN VO VDVO [- CO CO CO VQ VO VO -4 t—co O N O N L O o\t>--4rH rH  -=J- V D  H  c—co vo LCN co] cu -=t- vo LTN L O | C A co o  c— t—co ON 0"  rH  CO  r— O -4- -4-  cy O J O N 3 H C 0  rH  LOVO O-cO O N O H O J CO-4- U N CMOJ OJ OJ OJ ro ro ro co ro co o -4j CM, CO CO VOCO C—CO OJ H  C - C O CT\ C — O N V O  !^-V0  rH i-H  O N ^ ^ H  rH  3^3 3  OJCOHOJCO  H  CO VO C O  C— r-j VO VO r-j  H  L<\^*  O  LTN LTN LTN *t— LO VO t—CO LTN LCNLTN co ro -3- -4- LTNCO t- CO ON ON D - V O  CO VO VO LOVO-4-CO tTNJ- VO -3" VO t - ^ f LTNVO C*"- LCN L O  OJ OJ rH  C—  CO  CO  -3* VO ONt^t>-CO t-CO  O co  OJ  rH  r-i rH LT\ O LTN rH  CO  O N LTN O  t— V0  co  CM  t — C O O N O H CM oOJ- L O J V Q t—co O N O H CM rH H H H H H H H H H C M CM OJ OJ CM VO CO o o ON H H o I A O C O H co g ONCO rH rH H H H H  LCNVO  vo CO CO CO -4-CO t—CO CO VO t— LTNCO CO VD LTN  C—  t~- t — H  H  H  r|  t~-  LCN  t - V O C O t^--4" f-VO t— t— t— CO ONVO C—  O VO C J C O  OJ ON a  H  ON C O ON ONCO  t— C—VO - J -  VO  LTN LTN  O N C— O  g  co CO co O  rH rH  rH  co  O N C O -=)- r-  ON00  g  C A g ONLTN|^-  a •  • t—  S  H  VO P)LT\  a rH CQ  ON ON O  rH  rH H  O  H  ON 00  vo LCN  rH  vo g t-vo Hco t - 0 0 t~-co CM  O  O  C O C O C~-CO  -3- O  LO  LCN LCN L O  C O C O C O  CM  LTN  D-  rH rH  O  r*H  t—vo CO t—vo  C-— L O , ^ c — V O  co  LO, O N O N  CO  O N LCN LTN O N CO CO VO  ON^t  CO CO CO  VO  VO  H V OL O t^-LCN  co rOCM g  t-VQ  ONCOCO  o r-i  3 Sample No.  U " N _ d - VO  VO VO CO  rH  ON ON -4-  Table 3 ~ Distance function matrices for f i r s t and second runs ( Jp- ) o  O N C—  tr—vo o D— LTN] ro LCN  ONLOVOVO  LCN O CM  co  "—I rH ONVO C O O f  VO rH ONVO LCN LCN CO VO C— f - V OcO ON  C O H  a\co CO t— O C—CO  Run 1  - 12,3 CA H CM .3- vo VO- t>»co O N O H CVJ OOj-4- L T \ V O t — C O j C A o H O J oo|--t LTNVO t — C O ON O H OJ OO OO-t -4- - 4 - 4 - 4 - -4I A U N L T , LP. 1 A I A L P , LTN L P ,LfNVO VO VO VO H CM CM CM CVJ OO O* OO  ON-4 H  J  OO VO  VO LTN - N - - J - -4- 4 - C— ONVO r-j - 4  $  VO - 4 O VO VO t—r-j CO CO ON  H VO  .4- - 4  I—•I  co vo  t—CO VO O LTN  ON LP,  CO CO VO VO VO h  IT— LP,  ocoqcgl  ON ON L P , _ * VO vO ON H H  O CVJ H  VO  LTN LTN  LfN t>-_±  -4 LP,  CO CO VO OOVO  OO LTN  VO VO L T \ _ t _fr OOVO 00  t"-VO-4--4- -4-VD  OO OO^rt I —CO VO r-j s  -4-  C— O - V O  vo O N co co o  0 O - 4 LfN CO CO t —  CO -4-  •• r— t VO LP, tr—-4- -4— t  VO  tr-  0>  r—vo oo  .4  L T N C O vo  OO LP  LP,  -4"  CM -4"  f-  Q ON ON o co r-vo vo — 'I rH  H -4" O -4\ ON OO  O-CO  f-VO  ON g  r-j c o o o O N - 4 - 4 O L P , ON  CO  rA  OO  oo  VO CM -4 OJ  cv! H OJ  3 ON  t— t— L P , _ t -4H  H  q oo v o  1  00  C O O N V O O N c--  vo r — t L P , L P , rr LP, t— LP, LP.  f~ t—vo 00  r-J  O O V O C O -4"  C— f — C O t - v o LP, l>-CO co o  oo J  o rH  Sample No,  3  t-_t  c— o VO  O-CO COVO - 4  0 O - 4 L P . _ t vO  VO LfN-4* -4-  oo_t  on-4-  CVI - 4 ' -4-  - 4 -4" OO CVJ  rH c- C— OO LfN 0 O - 4 - - 4  -4-  LP.-4- - 4 ' o r L A CVI  OOVO LfN_t _ 4 - 4 -  LP OJ t— LT\ C~  VO f— OO  L P , _ t LP. LP -4" VO t-—  LT\ CV! CO L P . - 4 r O - 4 -  co - 4 '  f-co  LfN_f  oo oo  _j- _ t CO LfN H CM  00  C—^J- - 4  ON CO H O H H  O VO r-J  00  LP, f - ON  3D ON LP,  -4" C O V O V O C O 4 ' I>-  i—S  rH r-J  ONCO  X>  O N L P , Hco  VO  -4" VO -4" LP< LfJ  VO VO  t-  Table 3 (Cont.)  t— LPvVO CO Of  CACO  H  CVJ t~  oo  - J - - 4 CO -4" CM LP, [—-VO C O -4" _j- t*— LP*co L P _t _t c o -4-  -4-  oo  o\ H^o'-4' co vo vo  LP* V O LP, C O  — t r— t i-j ON  tr~ t r -  0 0 LfNVO LfN t - V D VO VO CO t - V O  LO _ t OO t™ LTN-4* -4- _ t OO LP.-4-  C O V O C O LP  LP, D— LfN  00H  t— O N J - V Ot— r-j t—CO C O  ON  O N LTN  LP. t- <~> V O CVj LP, H  f— C— H V O V O O J  IT— t * - - *  oo  LPVOO  VO LfN C— LfNVO  LP,  CO-4"  io oco  I—' H LfN O C — C O H rH _ * CO CO I— O  VO VO VO  OO J - _ t LP. LP>-4-  f— H CO CO CO CO LfN VO LTN VO LP, _-- LP, LP. LP VO VO t— -4- -4-  LfN LTNVO O H  C— t— LfN 0O_4-  oo ro cvl _ t  C - ONVO -4" L P . - 4 VO t — V O t -  oo ON LTN t— tr-|vo oo LTN CM O  LfN i  ON  3 -4  oo-4-  irvvo CO LTN  LfN t — L T \ C O t—  H  VO LfN t — _ t - 4 ' L f N O N V O V O CO  •-4 O K VO -4  c— c— t>-co vo - 4  t ~ " L f N O | J - ON LfN CO  O LP,  -4-  vo oo L P , oo oo  H ON C— C— LfN VO vO ON H CO H V O O t— O N h-|vO _ t C O LfN LP -4rH  VO LTN  CM LTN H LTN  ON  t-  rH  LTNONCO t— O -4" CO VO trH  O VO  co LTN  H  4 - C O 0 - LP.  -4  o vo co  Run_g.  VO  -124a  Table 4 - Results of third test, Run 3.  o  o  •rl  ri <rf  -P  o «S O  S-H  a o  •P  «H  •rl  CO CO  aj rH  o  K Pi m  Species A. esculenta A. tenuifolia  0.9 0.9  P cd  0 =» u u  0)  •rl  H  fl fl 3 IH «5 fl H  •P U -P fl <U <H •—f CO •rl -P  <D H  10 +3  H PH  n  2 2  1 2  0 1  3  <+H  4  4  -B)p <d P <U •P M aJ  d  0  •H CO  A •H 24 0  to O CO O  s o 2- pq  1 5  V M aj O co <D K  0 "8  «H  O JD H CO  a) _ •P P<  •rl O  <u  •H  rl P  ^ o _  -P  fl aj  •rl U O 3 -P O OS Si o P O l_,O B  0 7  1  . 2 4 5 B  I  10  23 1 5  -p a  _  CO  17 37  7  11  5  18  fistulosa  0.9  l 5  1 5  2 5  5 5  _8 15  A. praelonga  0.9  0 5  0  3 5  I  _8 15  A.  A. marginata  0.9  A. nana  0.9  A. erispa  0.9  5 4*9  1 _2 12  3 5  _4 25  0.7  6  Unknown from Adak  0.4  l  Species Not Covered  0.95  marginata/ tenuifolia Intermediates  0.8  8 T4"  2 3  marginata/ nana Intermediates  0.8  1 2  3 7  30  11 3*  14  W  0 2  1 1  5  14  _7 22  _7 22 0 5  Unknown from Sitka  Total  1 2  5  0  1 1  0 5  l 5  5 5  J> 15  _7 11  7 ii  2 5  2 5 l  l 1 3  11  11 20  _4 17  20  Jj.  0 1  10  Z  _4 10  10  Ji 33  _2 20  58 [203  11 33  27 27  9_ 2S3  In each c e l l , the numerator indicates the number of misclassifications and the denominator the total number of individuals in the c e l l .  •125A  B  a  a)  o  O H <1>  m  H  •top •H  Table 5 - Results of Run 4 (in Base .10 Logarithms.) E D H K HI OI <M| »H|  col co! a)  d u Pi  ail -P CD  SI  •H  a5  p. CO  HI O  •H  SI  o <D com  •HI  fl>!  OH-P o  < 32  ..20 .15 .03 .91 l.li 2.13 -.59 • 51 1.04 -.43  -.11  -.04 0.0  30  64  .03 -.06 -.02 -.16 -.22 - . 4 0 .32 .59 • 30 .82 1.70 1.59 -.80 -.90 .40 .47 1.05 1.13 -.26 .04 -.25 - . 4 8 -.14 ^1 15.9 o.o  39.2 11.5 0.0  -.26 -.26 -.38 • 33 .65 1.42 -.88 .16 .90 -.38 -.45 -.41  27.8 8.8 12.1 0.0  67  .Sample Size ' 40_ 59 79  Means of each variate -.03 -.10 -.29 -.29 -.03 -.17 -.18 -.49 -.05 -.15 -.37 .04 .55 .48 .56 .94 .80 .72 • 53 1.50 1.76 1.59 1.35 -.96 -.62 -.91 -.71 .08 -.08 .50 .10 1.07 1.30 .77 1.17 -.53 -.22 .11 -.11 -.52 -.10 -.18 -.39 -.23 - 3 3 Distance squared 39.1 36.6 51.0 17.0 25.2 27.7 15.7 35.6 31.2 5-5 18.8 20.3 0.0 25.9 35.1 0.0 14.2 0.0  78 -.01 -.02 -.07 .71 1.42 1.15 -.52 .85 1.19 .29 -.25  -21  matrix 33.5 '46.9 8.1 27-3 13.0 41.5 9.8 32.6 19.5 53-5 28.0 42.8 32.2 45.0 0.0 20.4 0.0  97  94  -.27 -.37 -.33 -.29 -.31 -.26 .64 .58 .86 .63 1.66 1.69 -.97 -1.01 .22 .15 • 97 1.02 -.25 -.33 -.32 -.27 -.46  36.6 17.7 23.4 9.7 18.5 18.5 17.2 15.3 40.5 0.0  40.8  21.2 34.4  13.9 23.3 23.3 20.5 23.1 37.0 5.1 0.0  79 -.23 -.25 -.34 .79 .77 1.74 -.95 .15 .96  -.41 -.39 -.36 30.7 14.4 21.6 7.5 17.2 25.6 25.3 12.4 37.0 5.8 8.6 0.0  -126Table 5 (Cont.) Variate: 8 .125  ,129  10  11  Standard deviation pooled over a l l species • 125 1 .269 J .2381 .23JT".152| .187*1 .2031 .225 .113 Grand mean 0 1  .012 .017  Pooled within-specles covariance matrix .008 .011 .012 .016 .012 .014  .003  .004  .007 .003 .013  i.6o  1.16 l.li  234.2 168.1  Between species covariance .matrix • 35 .62 1.08 1.09 1.36 .58 .59 .20 .05 • 72 .98  110.8 86.9  Eigenvalues 21.1  •127APPENDXX III. — Geographical  distributions.  Table 1. — New localities for the species of Alaria. Ajpjgrox. sosition  Locality  Apjprox. Locallto  position  A. ANGUSTA South Kurile Islands. Toshimoi, Etorofu I.  44N 147E  Hokkaido •Ainukappu Rocks. Akkeshi. *Daikokujima  •'•  42K 144E Eoinashiri, Eidaks. 43N IkkE Tengu iwa, nr. Chiexabetsu 42N 145E  44N 14JE  4lN 143E 42N 142E  Samukawa Shoya, Hidaka.  41N 1.40E 42N 143E  42N 1U3E  Tsukushiz-a, Kushiro  A. CRASSIFOLIA Hokkaido *Erimo Mombetsu, Hidaka. Nedenai, Oshima, Okoshimura, Hidaka. Honshu 39N 140E Iwazaki, Rikuzen. Kamanasaki, MiyagiKosode, Kuji Wan, Iwate. 40N 14IE 40N 141E Nakanomura, Iwate.  Same, Mutsu.  40N  Shiriya, Mutsu., Tutwatashi, Miyagi,  41N 141E  54N 52N 52N 52N  North I . , Adak I. l64w Sarana Bay, Attu I. 173E *Trappers Cove, Adak 173E Unalaska 1 . 173E *Zeto P t . , Adak I.  51N I76W  60N  172W  lUlE  38N 1 4 1 E  A. CRISPA Aleutian Islands *Cape Agagdak, Adak I. *Cape Sarichef *Casco Bay, Attu I. Chichagof P t . , Attu. *Murder P t . . Attu I.  52N  I 7 6 W  Sea St. Matthew I.  53H 173E  51N 176W 53H 167W 51N 176W  -128 Table 1.  (Contd.)  Locality  Approx. Locality  Approx. DositiOn  4-9N 48N  OW Portmazieres lW Quimper.  kkn  4 8 N 4W  3W  51N 58N  5W 3W  51N 55N 50N  9W 5W 5W  josition  A. ESCULENTA France Arromanches, Brest, Point de Primal, n. of Mordaix. B r i t i s h Isles Aber Mawr, Pembroke. A k e r g i l l , Caithness. Banfro. Bantry, Co. Cork. Campbell town, A r g y l l . *Cape Cornwall •Coosvaddig, Co. Kerry, *Cor Eileann, A r g y l l . Cullen, Banffshire Drom,Co. Kerry, *Eddystone Rock •Farland Pt., Bute • F i f e Ness, F i l e y Brigg, Yorkshire Glenbarr, A r g y l l . Inner Fame I., North. •Helmsdale, Sutherland,  52N  10W  55N  5W 2W  57K 52N 50N 55N 56N 54N 55N 55N 58N  10W 4w 4w  6ON 6ON  5E 5E  6ON  5E 5E TE  2W OW 5W 1W 3W  Kinghorn, F i f e . Land's End, Cornwall, Lossiemouth, Mor. Malin Head Mbvllle, Co. Done. •Mull of Kintyre Portnahavn, I s l a y Portrush Port S t . Mary. Renoyle, Comiemara Rosneigr, Anglesea. Sanday, Orkney I s . Scapa Flow, Southend, A r g y l l * Trotternich, Skye Valentia •Wee Curabrae I.  56N 50N  571 55N  55N 55N  55N 55N 54N  53N 59N  58N 55N  3W 5W 3W 7W 7W 5W  6w 6w 6w 2W 3W 5W  51N 55N  6w low Uw  68N  ITE  TON  22E  TON  30E  57N  Norway Algeron, Bergen, Edonsiberg, nr. Bergen Flaese, Rogaland > F o l s v i k , Bergen, Haugesund •Kristiansund L i s t e r , Vest Adger •Maloy  59N  63N 58N 6IN  6w  62N 6lN  6W 6W  64N 65N 65N  22W 18W 13W  5W  Narvik, SandS, Nordland SorOy I . Stavanger Troaso. Trondheimsfjord Skalanes• Vardo .  ;  58N 5E 69N 19E 63N IDE  Faeroe Islands Muggeness, Holm.  0stero, G l i b r o , Sand$<  Thorshavn, Strt^m^. Videro, Vedoy  62N 62N  G r j a l n e s a,, Melrakkasljetta Holmanes, E s k l f j o r f t u r  66N l6w 65N 13W  6W 6w  Iceland •Akranes. Akureyri. B or gar f j or^> ur.  -129-  Table 1.  (Contd.)  Locality-  Approx. position  Locality  Greenland Claushavn Exkrementbrigten Husavik Fiskerness.  63K 51W  Issa, ArsukfJord Julianhaab• Karajak Kobbereinoen nr., Julianhaab  6lH  48W  6ON  46w  60N  46w  Canadian Arctic Koksoak R., Ungava  58N  68W Northumberland Sd.,. Queen's C h a n n e l  54N 60N  57W Hebron, 64W Strait of Belle I.  Labrador Hamilton Inlet Port Burwell.  58N 62W 5 IN  56W  45N  66W  Maritime Provinces and Quebec Cap a l'Aigle, P.Q. 47N Cape Forchu, nr. Yarmouth43N 4?N Channel, Nfd. kkN Grand Manan I . , N.B. 48N Matane, P.Q. MiqueIon I . U'TN 49N Mont Louis Bay, Gaspe  70W Peak's I . , N.S. 66w *Point Lepreau, N.B. 59W Sambro I . , N.S. 66W *nr. Sheet Ebr., N.S. 67W Torbay, Nfd. 56W  44N 62W 47N 52W  65W  New England Appledore I . , Bay of Shoals B r i s t o l , Me. Casco Bay, Me. 43N Eagle I . , Me. Eastport Me. 44N Hampton, N.H. U2N Harpswell, Me. Cape Ann, Mass. 42N  70W 67W 70W 70W  Magnolia, Mass. Matinic I . , Me. Mt. Desert I . , Me. Quoddy Head, Me. Rockport, Mass. Sequin I . , Me. (drift) Weymouth, Mass.  43N 68W 44N 68W  UUN 67W 1+2N 70W 42N 70W  Soviet Arctic Rogatschev Bay, Novaya Zemlye.  71N  52E Tjapka, Siberia Sosnovetz I . , White Sea  Aleutian Islands *Casco Bay, Attu I. Chichagof P t . , Attu I.  52N 173E M i l l Bay, Kodiak I. 52N 1 7 3 E  57N 152W  -130Table 1 (Contd.) Locality  Apjprox. position  Locality  Approx. position  A. FISTULOSA Gulf of Alaska •Seldovia-  59N 151W  Aleutian Islands Belkovsky I. •Cape Agagdak, Adak I. •Cape Sarichef, Unimak I.  55N 1.52W •Murder P t . , Attu 51N I76W Shumagin Is. 54N l6UW Zeto Pt-., Adak I.  52N 173E 55N l60¥ 5IN 176W  U6H 150E  Toshimoi, Etorofu I.  45N  ikjE  Kurile Islands Urup I. Hokkaido •Abashiri (drift)  A. GRANDIFOLIA Jan Mayen I.  TIN  8W  63N  51W  Sukkertoppen, Kangerdlugsuak  65N 6TN  53W 52W  58N 58N  62W McClellan Str. 68W Nanook Bay Strait of Belle I.  60N  64W  51N  56w  56N 135W Knight I., Yakutat. 59N 139W -"Western Channel, Sitka,  59N  Greenland Danmarks <j>, Hekla Havn Fiskerness Labrador Hebron. Koksoak R.  A. MARGINATA South-east Alaska Biorka I . , Sitka. Khantaak I . , Yakutat  5TN  139W 13 5W  British Columbia Aguilar Pt. Bartlett I. •Beacon H i l l Park Bch.  48N 125W l*-9N 126W 48N 123W  Deer I. Deer Island Reef Donaldson I .  50N 12TW 50N 12TW U8N 123W  \  -131Table 1 (Contd.) Locality *Big Bay, Spring I. Black Creek Black River *Brooks Peninsula Cape Sutil Catala I. Chanal Reef Darling River Kains I. Keogh R. Klucksiwi R. Larigara I. Lawton Pt. *Loran Station, Spring I. *MacLean I. Magdalena Pt. Maquinna Pt. Mills Peninsula Muir Creek *Ogden Pt. Breakwater *Otter Pt. Perez Rocks  Approx. position 50-N 127W 48N 125W  48K 124W 50W 12TW  50N 128W  49S 126W 54N 133W 49W 125W 50N 50N 50N  128W  127W 12?W  54N 133W 48N 125W 50N  127W  50N 12?W 48N 124W 49N 126W 48N 125W  Locality Elsenia Rock Estevan Pt. Flea Beach Garden I. •Glacier Pt. Hope I. Iron Mine Bay Plover I. Ralston I. Redfern I. •Sandstone Creek Sharp Pt. Sheringhara Pt. Striae Pt. Tremayne Bay Valdez I. Mazzaredo Is.  Approx. position 50K 127W  49N 126W  48K 123W 50N 127V7 48N 124W 50K 127W 48K 123W 50K  127W  50K 127W 48K 124W 49K 127W 48K 123W 54K 132W 50K 127W  54H 132W  48N 123W  48N 123W 48N 123W 49N 127W  Washington Crescent Beach *Eagle Pt. Fidalgo I.  Boiler Bay, Lincoln Co. Brookings Charleston, Coos Co. Cape Blanco Harris Beach  Goose I. 48W 123W *Mt. Dallas Beach 48N 122W Waadah I. Salmon Bank  48N 122W 48N 123W  48K 124W 48N 122W  42K 124W 43N 124W 42K 124W  *Nr. Newport North Bay, Cape Arago Seal Rock Sunset Beach Yaqulnna Head  44N 124W 43N 124W 44N 124W  36K 121W 37N 122W 36N 121W 36K 121W  San Simeon Santa Cruz Ten Mi. S Pt. Sur Trinidad  35N  121W  40N  124W  44N 124W  California Carmel R. Greater Farallon I. Mission Pt. Pacific Grove  36N 122W 36N 121W  -132Table 1 (Contd.) Locality  Approx. position  Locality  Approx. position  A. NANA South-east Alaska Biorka I . , Sitka. Western Channel  56N 135W •Ocean Cape, Yakutat 57N 135W  59N 139W  British Columbia Amphitrite Point Bartlett I. *Big Bay, Spring I. Blingkhorn Pen. •Botany Beach, nr. Port Renfrew Cape Sutil Deadman I. Donaldson I. •Glacier Pt. Kains I. •Kelsey Bay  U8N 50N 50N 50N 1+8N 50N U8N U8N 50N 50N  •Loran Station, Spring I . Mills Peninsula •Otter Point .Plover I. Sharp Point 12l*W Race Rocks 128W Sombrio Pt. Wouwer I . 123W 124W 128W 12 5W  125W 126W 127W 126W  50N 127W i+9K 12 5W koH 123W 50N 127W >9N 126w 1+8N 123W U8N 12*W IIS! 12 5W  Washington Waadah I.  Crescent Beach  1+8N 12l»W  California Moss Beach  37N 122W Pescadero Pt.  36N 121W  A. OCHOTENSIS Chir.ihasan, Saghalien  49N  ihkE  A. PARADISEA none A. PRAELONGA Aleutian Islands •Cape Agagdak, Adak I . •Cape Sarichef, Unimak I. Kamchatka Ma syria  52N 173E North Arm, Adak I. 5*t-N 16UW Unalaska, I.  51N I76W 53N.l66w  -133Table 1 (Gontd.) Approx. position  Approx.  Locality South Kurile Islands Chikohai, Etorofu I.  Chinomichi, Kunashiri  1+4N 1U6E  Kiritup, Kushiro Kombumori, Kushiro Kushiro Hbr. Okoshi, Hidaka Tomoshiri, Nemuro  1+3N Ik 5E kzs I U E  Hokkaido Akkeshi •Daikokujima Hamanasaka, Kushiro Hanasaki, Nemuro  43N IkkE  4 2 N 145E 43N 145E  1+2N l U E  UlN lU_E  1+3N 11+5E  A. PYLAII Labrador N. shore Newfoundland Maritime Provinces Digby, N.S.  kkS  66w  A. TAENIATA Bering Sea 57N  St. Paul I.  170W  Aleutian Islands •Finger Bay, Adak I.  51N 176W  Gulf of Alaska Cape Chiniak, Kodiak I. •English Bay Homer  57N 152W M i l l Bay, Kodiak I. 59N 151W •Pasagshak Pt. 59N 151W •Seldovia Point  South-east Alaska •Halibut P t . ,  Sitka  57N 135W  A. TENUIFOLIA Norwa\ Trondheimsfjord  63N 10E  57N 152W 57N 152W 59N 151W  -134Table 1 (Contd.) Locality  Approx. position  Locality  Approx. position  Spitzbergen Bootsund Iceland EskifjorW Hognarsdarif  651ft •13W  Reydarf j o r V i r Sfcatnesf jor^ur  65N  13W  Greenland Maneetsok, Arsukfjord  6lN  48W  A l e u t i a n Islands nr. Atka I .  52N 174W *Chichagof Bay, A t t u  52N 173E  58N 134W  5TN 134W  South-east Alaska Juneau Port Augustine  Sitka  B r i t i s h Columbia Albert Head Balaclava I . Blinkhorn Pen. Brockton Point Campbell River Cape Lazo Chatham Channel Dodds Narrows Gordon Head Kelsey Bay Keogh R . , south of Klucksiwi R.  48N 123W  124W 12 5W 12 5W 126W 12 3W 127W 123W 125W 123W 123W 127W 127W 126W  Malaspina Narrows Maude I . Miracle Beach N e v i l l e Point •Ogden Pt. Breakwater Port McNiel Race Rocks Rosebush I . Sechelt Narrows Smythe Head Sointula 50N 127W Ten Mile Point Walker Point  50N 50N 49N 50N  48N 122W •Friday Harbor 46N 122W •Turn Point 48N 122W Whidbey I .  48N 12 3W  50N 50N 49N 50N 49N 50N 49N  127W 126W 123W 12 5W 124W 126W 124W 48N 12 3W 50N 126W  48N  50N  48N  50N 49N 48N  50N  48N  50N  Washington Brown I . Davis Bay, Lopez I . Goose I .  48N 122W 48N 122W  The approximate positions above are given to degree only, of latitude and longitude. The degree given is the degree which is numerically smaller than the complete bearing, and so is not necessarily the nearest one. Localities, from which samples were used in the s t a t i s t i c a l analysis, are -marked with an asterisk.  -135Table 2 - List of localities cited elsewhere than in Appendix III, Table 1. Locality o  Alesund Abashiri Aberdeen Adak I. Agattu I . Ainukappu Rocks Airoppu (Saghalien) Albert Head Amaknak I. Ambetsu (Saghalien) Amphitrite Pt. Anglesea Araito I. Argenton (France) Arimoe (Etorofu I.) Asa I. Atka I. Atoiya (Kunashiri I.) Attu I.  Approx. position* . Locality  62N 44N  5?N  6E IkkE  2W  51N 176W 52N  173E  43N 145E 4 6 N 143E 48N 123W  53N 167W 50W 142E 48N 12 5W 4W 53N  50N 155E  45N  147E  52N 174W  44N l 4 6 E 52H  Cape Sarichef (Unirnak I) Cape Ustrechni Carmel Bay Casco Bay (Attu I.) Cherbourg Chiboi (Shikotan I.) Chicagof Bay (Attu I.) Chikohai (Etorofu I.) Chinomichi (Kunashiri I.) Chitose-wan (Paramushiru I.) Commander Is. Cook Inlet Coosvaddig (nr. Dingle, Ireland) Cordova Cor Eileann Crescent City  173E  Daikoku-Jima Deer I. Beacon H i l l Park Beach (Victoria)48N 123W Dobuki (Saghalien) Bering I. 55N I67E Douglas Bering Strait 65N I69W DrankinskI 55N Berwick-on-Tweed 2W Dshukdshamdram Bettobu (Etorofu I.) 45N 1 4 8 E Dunbar Big Bay (Spring I.) 50N 127W Botany Beach (nr. Eagle Point 48N 124W Eddystone Rock Port Renfrew) 48N Brittany 4W English Bay 123W English Channel Brockton Point 49N 42N 124W Esquimalt Brookings 127W Brooks Peninsula Etorofu I. 50N Broughton Bay (Shimushiru L^7N 1 5 2 E Etorofu Strait Eyrarbakki Busumisaki (Etorofu ImCape Cape Cape Cape Cape Cape Cape Cape  52N 176W Agagdak (Adak' I.) Atoiya (Kunashiri I.) 44N 1 4 6 E 72N l8w Bismark 50N 5W Cornwall 4 l N 143E Erimo Hope (Greenland) Nichta 45N 142E Notoro  Fairhaven (Spitzbergen) Fanafjord (Nr. Bergen) Farland Point Fife Ness Finger Bay (Adak I.) Foehr Fort Ross Friday Harbor Funkakoshi (Saghallen)  Approx. Position* 54N l64w 36N 122W 5 2 N  173E  49N 1W 43N I46E  52N  I73E  44N  I46E  45N 148E  55N 1 6 7 E 152W  60N  52N 10W 60N 1 4 5W 56N 5W 4lN  124W  43N 145E  50N  127W  143E 58N I34w  46N  56N  2W  48N 123W  50N  4w  49N  4w  59N 151W  48N  123W  45N 148E 45N 149E 63N  20W  60N 5E 55N 4W 56N 2W 5IN I76W 54N 8E 38N 123W 46N 123W 46N 143E  -136 Table 2 (Cont.) Approx. . ]Position* L o c a l i t y  Locality G l a c i e r Bay G l a c i e r Point Godhavn Golden Gate Great Shantar  Approx. Position*  Klueksiwi R. Kodiak I. 69N 53W Koldeway I. 37N 122W Koljushin 55N 138E Korsfjord (Nr. Bergen) Kosikagawa (Shumushu I.) lUOE UlN Kraan Cove 57N 135W fc9N15^E Kristiansund (Norway) 57N 153W Kukak Bay 58N 3W Kunashiri I . Kurosaki (Paramushiru I.) 4 2 N lk2E U6N lUlE. Kvarno I. Ij-W Kynance Cove (Cornwall) 53N 50N 127W Kyuquot 58N 136W  1+8N 123W I.  Hakodate Halibut Point (Sitka) Harumukotan I. Harvester I. " Helmsdale Hidaka Hishitoma- (.Saghalien) Holyhead Hope I» Idashibenai (Kunashiri Ikerasak (Greenland) I l l u l u k Bay I r i b u t i (Etorofu I.) IsfJorden I s l e of Man J s l e of May  I.)  70N  I46E 51W  53N 167W lk8E 7 8 N 15E  56N  2W  71N  Kaiba-to (Saghalien) Kaikyo-to (Saghalien) Kains I. ICakamusbetsu (Paramushiru I.) Kamschatka Kamuiwakka (Etorofu I.) Kangerlugsuak (Greenland) Karaginsk I. K a r l s k a l i (Iceland) Karluk Kataoka (Shumushu I.) Keku Inlet (S.E. Alaska) Kelsey Bay Kenai Peninsula Ketchikan Ketoi I. King's Cove (Cornwall Kitazima (llshishiru I.  kGs l U l E  6W  50N 126W 58N 123W  58N 13^-w 48N ikhE  50H 128W 155E 160E ikSE  53W 161+E  13W 15UW 501 1 5 6 E  50N 126W 50N 150W 55N 131W  VfN 152E l+TN 152E  5QN 156E 49N 125W 63N 7E 58N 154w  kkH ik6s 50N 15 5E U8N 125W 1+9N 5W 50N 127W 37N 122W  kSs I.)  4w Maaso (Norway)  Jan Mayen I. Johnstone S t r a i t Juan de Fuca S t r a i t Juneau  50N 59N 45N 68N 59N 65N 57N  Land's End Le Conquet (France) Lodingen Loran Station (Spring  50N 127W 57N.. 154W 7N 18W 67N 17^E 60N 5E  Maclean I . Maloy Mamgabai Mandal Matlsett Narrows Matsuwa I. Medni I. Millport Minamiura (Araito I.) Minamiura (Ketoi I.) Minami-ura (Urup I.) Minamizima (Ushishiru I.) Minnesota Reef Miquelon Mi'shima (Urup I.) Miyagi Prefecture (Honshu) Moiretomori (Saghalien) . Monterey Bay. Mossel Bay Moyoro (Etorofu I.) Mt. Dallas Beach (San Juan I.) Muir Point Mull of Kintyre  kv  68N 1 6 E 50N 127W  50N 127W 62N 5E  58N TE 1+9N 123W  k&E 153E  55N 4W 50N 15 5E  U7N 152E ->8S? 122W 57N 56W k6u 1 5 0 E  38N 141E 4 b N 141E 36N 121W U5N lk&E  J+8N 123W U8N 123W 55N 5W  -137-  Locality Murakami (Paramushiru I . ) Murder Point (Attu) Muroran Mushiru I . Mushiru S t r a i t Naibo (Etorofu I . ) Navoshi (Saghalien) Neah Bay Nemo-wan (Onnekotan I.) Nemuro Newfoundland Newport (Oregon) Nijo-iwa Notoro Peninsula  Table 2 (Cent.) Approx. Position* Locality  50N 52N 42N k&S 48N  Approx. Position*  I56E 4&V 140E 153E 153E  Race Rocks Raunfjord (Nr. Bergen) Reykjavik R e y W f j o r W (Iceland) Roche Harbour (Wash.) Roecoff (Prance) 45N 147E Rabetsu (Etorofu I.) 49N 142E Ryrkarpiya U8N 124W 49N 154E St. Dimon (Faeroe I.)  43N Ik5E  St. Ives  48l 56W St. Lawrence Bay kkTS 124W St. Lawrence I . 45N 142E St. Michael's Mount 45N 142E St. P i e r r e Sandstone Creek k&S 123W San Juan I . Oak Bay 59N 139W S c i l l y I s . Ocean Cape (Yakutat) 46N 143E Sekinezaki (Araito I.) Ochibo (Saghalien) Seldovia Point Odomari (Onnekotan I.) Ogden Point Breakwater Seley 48N 123W Shakotan (Shikotan I.) (Victoria) Shana Gun (Etorofu I.) Okaboi (Etorofu I.) Oma (Honshu) 41N 140E Sheet Harbour Onnekotan I . 49N 1U3E Shikotan I. Orca (Alaska) 60N 145W Shimofuro (Honshu) Otter Point U8N 123W Shimushiru I. Ouessant U8N 5W Shiranushi (Saghalien) Ozernaya 47N 162E Shumushu I. Sitka Pankara (Kamschatka) 58W 162E S k a i l l (Orkney Is.) Paramushiru I . 50N 155E Skansbay (Spitzbergen) Pasagshak Pt. (Kodiak I.) 57N 152W Skaulefjord (Faeroe Is.) 49N 126W Sooke Perez Rocks 36N 121W Spitzbergen Pescadero Pt. (Cal.) 57N 2W Spring I . Peterhead Storabucht (Koldeway I.) Petropavlovsk 57N 158E Stonehaven (Kamschatka) 50N 142E S t r a i t of Georgia Pilevo (Saghalien) Striae Point P i t l e k a y (USSR) 50N 4W Suppertoppen (Greenland) Plymouth (UK) 45N 66W Surlbachi (Paramushiru I.) Point jjepreau (N.B.) 36N 121W S y l t Point Sur Point Aulon (Cal.) 48N 124W Table I. Port Renfrew k&S 124W Tenjin Iwa (Shumushu I.) P o s t e l s i a Point Tepsikova Reef Preobrajenuja Bay (USSR) 60N 147W T i b o i (Shikotan I.) Prince William Sound 47N 122W Puget Sound 50N 127W Pulteney Point  48N 123W 60N 5W 64N 22W 69N 13W 48N 123W 48N 3W 45N 147E 68N 179W  50N 5W  65N 171W 63N 170W 50N , 5W U6N 56W 48N 124W 48N 123W 4 9 N 6w ?0N 155E 59N 151W 64N 13W 43N 146E 45N 147E 44N 62W 43N 146E 4IN 141E 47N 152E 46N 141E 5°N 156E 57N 135W  48N 123W 78N 15E 50N 127W 5°N 2W 4-9N 123W 5^N 132W 66N 53W 50N 1 5 5 E 55N 8E  5IN 127W 43N 146E  -138-  Locality Titose (Paramushiru I.) Tokotari (Urup I.) Toshirari (Etorofu I.) Trappers Cove (Adak I.) Treureriberg Bay (Spitzbergen) Turn Point Tyulen'oi Rock Uebetsu (Urup I.) Uebetsu (Etorofu I. Ukhidnoi Reef Umatilla Reef Unalaska. Bay Unalaska I. Unga (Alaska) Urup I.  Table 2 (Coat.) Approx. Position* Locality 45N 149E lk8& 51K lT6w /  Approx. Position* 152E  Ushiro I . Uyak Bay  153W  Vancouver I. Victoria  50N 126W 58N 123W  55N  167W 55N 160W  122W 56N 132W  48N  53N  46N 150E  4w  57W 135W  v  59N 139W  Yakutat Yamato Wan (Matsuwa I.) Yessa (Urup 1.) Zeto Point (Adak I.)  *  48N 123W  48N 122W Waldron I. 60N 1T0E Wee Cumbrae I. (Scotland) Western Channel (Sitka) kkE 147E Whale Sound (Can.) 48N 124W Whidbey I. (Wash.) 53N l6?W Wrangell  k&B153E 45K IkkE .  51N 176W  The approximate position is given to degree only, of latitude and longitude. The degree given is the numerically smaller one and not necessarily the closest.  -139 (to face 140) -  Figure 1. A. B. C. D. E. F.  — — —  --  --  G.  H. J. K. L.  — —  --  Illustration of morphological terminology.  lamina. rachis) stipe ) holdfast. blade. midrib. transition zone. sporophyll. sorus. petiole. hapteron. t  r  u  n  k  #  The numbers correspond to the measurements defined on page 5«  -141 (to face 142)-  Figure 2.  Patterns of growth in Alaria esculenta.  The illustration shows a! plant studied between May 10 and June 5, 196l. The outline shows the form of the plant and the position of marks on June 5. The arrows^Indicate the movement of each of these marks relative to the mark proximal to i t . In this way the net growth between marks is shown, during the period between May 10 and June 5-  -143 (to face  Figure 3 '  lkk)~  Rates of growth in three Alaria species.  The. growth rates of the trunk in millimeters per day are shown below the line, and the growth rates of the lamina in centimeters per day above the line. The measurements were made on two year classes during the same year, not on the same individuals throughout two years. The sample size for each reading is given by the number adjacent to each point plotted. In cases where the number is small, the reading given was often supported by minimum values obtained from plants which were too damaged to provide a precise reading.  1 1 1 1 1 1 1 1  1  1 1 1 1 1 1*  6  2 / \  : >  ^  y  \  1  7  co  14  o 2:  1 —1—1—1—1—1—1— ( 1 1 1 1 1 1 1 1 1  M  9 -  13  8  7  14  5  14  *  11  - - - - - - - * ~ - ^ U ^  V~^~7  7  CO  >  1  >  — Z >  O Cfl  1  1  1  1  —1—1—1—f- M  1  i-  1  1  M  1  1  l  i  l  1  1  1  1  I  1 1 1 1 1 1 1 1 1  1  1  (  1  1  1  1  1  1  1  1  1  l  1A  •  1  •(=-  I  r-  J>. CO  -  14  2  /  \  1  * * 1. 5  .>  c z TV"  NANA  MARGINATA  4v  -  \  CO  _  *  \  ,-"14 V  .  "ii  5  2 '"-<•  0 z  , *i 1  1  V  TENUIFOLIA  u  1  1  1  1  1  1  1  1  1  1  1  1  1  1  J—  1  MONTH  1  SUPERIMPOSED 1  1  1  i  i  i  1  1  1  1  i  1  .  i  l  .  -145 (to face 146)-  Figure 4.  Type localities and geographic regions. (Circumpolar projection of the Northern Hemisphere.)  The arabic numerals which identify the type localities correspond to the numbers of the names listed on pages 19 f f . The geographical regions, under which localities are listed on pages 45 f f . , are identified by roman numerals. I. II. III. IV. V. VI. VII. VIII. IX. X. XI. XII. XIII. XIV. XV. XVI. XVII. XVIII. XIX. XX. XXI. XXII. XXIII. XXIV. XXV. XXVI. XXVII. XXVIII. XXIX.  Honshu. Hokkaido„ Saghalien„ South Kurile Islands. Middle Kurile Islands. North Kurile Islands. Ochotsk Sea. Kamschatka. Commander Islands. Aleutian Islands. Bering Sea. Gulf of Alaska. Southeast Alaska. British Columbia. Washington. Oregon. California. Soviet Arctic. Spitsbergen. Norway. British Isles. France. Faeroe Islands. Iceland. Greenland. Labrador. Canadian Maritime Provinces. New England. Canadian Arctic.  -147 (to face 148)-  F i g u r e 5. Some presence-and-absence factors in the environment of Alaria. Data taken from Isakov, 195 3..  y  -149 (to face 150)-  Figure 6.  Oceanographic features at the southern limit of Alaria marginata in California. Monthly means of daily observations -of salinity and sea temperature. A. B. C. D.  Pacific Grove, 1960-61. Hunters Point, San Francisco, 1955-56. La Jolla (surface), I96O-61. A, B, and C superimposed on the same coordinates.  Data taken from Anonymous, 1962.  10  15  T°C  -15020  10  15  CO CO"  CN. CO  CO  / *\ /  o. CO  \  /  \  \  i  \  \  c0 o  cV  \  *  \ \ \ \  09. CN  \  \  \  \ CM'  V m  3-  ro. CO  \ \  \  \  -151 (to face 152)-  Figure 7» Oceanographic features at the southern limit of Alaria esculenta in New England. Surface readings of salinity and sea temperature. A. B.  Portland lightship, monthly means, 1958-59. Passamaquoddy Bay, means of monthly observations,  C. D.  Buzzards lightship, monthly means, 1958-59. A, B, and C superimposed on the same coordinates.  1957-62.  A and C taken from Day, 1959, i960. B. provided by Lauzier, pers. comm. Data from Passamaquoddy Bay may be slightly outside New England and slightly Inside the geographical distribution of Alaria.  -153 (to face  154)-  Figure 8. Oceanographic features at the southern limit of Alaria esculenta in the English Channel. Salinity and sea temperature surface readings, at least one per month. A. B. C. D.  Station E l , 1961-62. 50°35' N . , 1 ° 4 0 ' W., 1951-53. 48°26' W., 5°17' W., 1951-54. A, B, and C superimposed on the same coordinates.  B and C taken from Anonymous, 1952, 1953, 1954. A provided hy Boalch, pers. comm.  -155 (to face 156)-  Figure 9« Oceanographic features at the southern limit of Alaria crassifolia in Japan. Salinity and sea temperature, ten day means of daily surface readings. A.  Hakodate, Hokkaido, 1961-62.  A provided "by Masaki, personal communication.  -15620  15  T * C  10  5  -157 (to face 158) Figure 10.  Lower limits of Alaria in the subtidal zone.  These data as available are not s t r i c t l y comparable and should only be considered in a general way. Data from the literature are actually deepest records; only the personal communcations are real lower l i m i t . A l l have been reduced to an arable number representing meters below low water mark for ease of presentation. This has been done even with Boergesen's record, which is stated to be "several fathoms", and Wilce's, which is the lower limit of an association stated to contain Alaria. The sources of these data are given by the roman numerals:: I. Enomoto, pers. comm. II. Setchell et. Gardner, 1903, p. 271. III. Neushul, pers. comm. IV. E . S. Zinova, 1929a, p. 86. V. Kjellman, 1883a, p. 217. VI. Svendsen, 1959, p..34. VII. Printz, 1926, p. 192. VIII. Kjellman, 1906, p. 11, and herbarium annotation. IX. Lund, 1959, p. 166. X. Jonsson, 1912, p. I83. XI. Boergesen, 1905, p. 754. XII. Kain (Jones), pers. comm. XIII, XIV. Boalch, pers. comm. . XV. Wilce, 1959, P. 51. XVI. A. R. A. Taylor, pers. comm.  -159 (to face l 6 o ' ) -  Figure H .  Beds of Alaria fistulosa in southeast Alaska. Representative localities taken from Frye, 1915, Tables XXXII, XXXIII.  -161 (to face 162)-  Figure 12.  Distribution of.Alaria in southwest England. Above:  Specimens collected before 1930 Probably drift Literature records only  Below:  Found i960 to 1962 Disappeared between 1950 and i960 (data provided by Boalch, pers. comm.)  ,  50  »  MILES  t  -I63 (to face 1 6 4 ) -  Figure 1 3 .  Results of the f i r s t run. The f i r s t two discriminant scores of each sample are plotted against one another, and the size of the squares are proportional to the third discriminant score.  •  6 SECOND  I  7  I  I  8 9 DISCRIMINANT  I  10  t  11 SCORE  1  12  I  13  I  14-  -165 (to face 166)-  Figure ik.  Frequency distribution of values. The distances plotted include a l l those of a l l samples in Run 2 except sample No. 60, as well as a l l samples in Run 1 which do not have common elements with samples in Run 2.  -166-  200 _»  T"  —  F R E Q U E N C Y  i  i  1  1  i  1-  i  lOO i  •  i  1  i  i_  1  1  O  r-  /  -167 (to face 168)-  Figure 15.  Results of the second run. The f i r s t two discriminant scores are plotted against one another for each sample, and the sizes of the squares are proportional to the third discriminant score. The Atlantic samples are stippled.  -168FIRST 15  14  13  12  11  DISCRIMINANT 10  9  8  7  6  SCORE 5  4  3  ~l69 (to face 170)-  Figure 1.6. Relationships "between the Atlantic and the Pacific samples. Every sample is plotted at the correct Jlf- from the mean discriminant functions of a l l Atlantic and a l l Pacific samples (+). The unlabelled Pacific samples are from the Aleutian Islands, and the unlabelled Atlantic samples are from the British Isles.  0  t  \/~D*  1  1  1  i V A L U E S  5  i  -171 (to face 172)-  Figure 17.  Results of the fourth run. The approximation of 12 dimensional hyperspace in-two dimensions of course-leaves much to be desired. This may be seen in the fact that thjj- lines between each species are often not proportional to the \Jlr values placed immediately beside each In a clockwise direction. The circles represent the following species: GF A TT P M TF F CP N EN EB ER  crassifolia — A. angjjgta. — A. taeniata — ETSalonga — A. marginata — A. tenuifolia -- 4. fistulosa — A.' crista ~ A. nana — A. esculenta. Norwegian samples. —• A. esculenta, samples from the British Isles. — 4- esculenta. samples from the rest of the world.  •173-  Figures 18 to 31. Geographical distributions of species, Key to symbols used; Type locality Samples used in s t a t i s t i c a l analysis Other localities  0 o  (drift) (drift)  A. A  -174 (to face 175)-  Figure 18.  The distribution of Alaria angusta.  -176.(to face 177)-  Figure 19.  The distribution of Alaria crassifolia.  -178  Figure 20.  (to face 179)-  The distribution of Alaria erispa.  -lBO (to face l 8 l ) -  Figure 21.  The distribution of Alaria esculenta. A, B, and C correspond to the localities of the oceanographic stations described in Figures 7 and 8.  -182 (to face 183)-  Figure 22.  The distribution of Alaria fistulosa.  -184 (to face lfl5')-  Figure 23.  The distribution of Alaria grandifolia.  -186 (to face 187)-  Figure 2k.  The distribution of Alaria, marginata. A, B, and G correspond to the localities of the oceanographic stations described in Figure 6.  -188.(to face 189)-  Figure 25.  The distribution of Alarja  -190 (to face 191)-  Figure 26.  The distribution of Alaria ochotensis.  - 192 i(to face 193 )-  Figure 27.  The distribution of Alarja paradisea.  -194 (to face 195)-  Figure 28. The distribution of Alaria praelonga.  -196 (to face 197)-  Figure 29.  The distribution of Alaria p y l a i i .  -198 (to face 199)-  Figure 30.  The distribution of Alaria taeniata.  - 200 (to face 201) -  Figure 31.  The distribution of Alaria tenuifolia.  1  -202-  Figures 32 to 4-5.  Illustrations of species.  A characteristic individual of a species is shown in the centre, of each i l l u s t r a t i o n . A trunk and a. sporophyll of an individual intermediate in form between the species in question and a species similar to i t are shown on the left hand side of the illustration. The trunk and sporophyll of an intermediate with another similar species are shown on the right hand side of the i l l u s t r a t i o n . The name of the species to which these intermediates show similarities is shown beside the trunk. Four species, Alaria grandifolia. A. ochotensis. A. paradisea. and A. p y l a i i were not collected by the writer, so he has no idea of how a "characteristic" individual would differ from the type specimen. So, in these cases, a characteristic individual other than the type specimen is not illustrated. :  Figure 3 2 .  Illustrations of Alaria angusta.  -204-  Figure 3 ^ .  Illustrations of Alaria crispa.  Figure 35.  I l l u s t r a t i o n s of A l a r i a esculenta  Figure 3 6 .  I l l u s t r a t i o n s of A l a r i a f i s t u l o s a .  -206-  Figure 38.  I l l u s t r a t i o n s of A l a r i a marginata.  -211-  Figure h}  9  I l l u s t r a t i o n s of A l a r i a p y l a i i .  igure hk.  I l l u s t r a t i o n s of A l a r i a t a e n i a t a .  Figure 4-5„  Illustrations of Alaria tenuifolia„  -217-  Figures 46 t o 59 • I l l u s t r a t i o n s of type specimens of species studied i n d e t a i l (Taken from photographs).  2 CM  Figure 46. The type specimen of Austria angusta  Figure 5 0 .  The type specimen of Alaria fistulosa.  Figure 5 1 .  The type specimen of Alaria grandifolia.  2 CM Figure 52.  The type specimen of Alaria marginata.  Figure 54.  The type specimen of Alaria ochotensis.  -222-  -224-  Figure 56„  The type specimen of Alaria praelonga„  2 CM  Figure 57. The type specimen of Alaria p y l a i i .  Figure 58. The type specimen of Alaria taeniata.  Figure 5 9 T h e t y p e specimen of Alaria tenuifolia.  -227Figures 60 and 6 l .  Illustrations of the type specimens of synonyms.  The number in each frame corresponds to the serial number given each name in the text (Table 1, p. 19) and in Appendix I ( l ) „ The scale in each frame is the„equivalent of 5 centimeters. Ho.  Name  Name  0  Alaria corrusata Miyabe  35.  A. laticosta Kjellm.  5.  Alaria curtipes Saunders -  36.  A. linearis Stroemf.  6.  Agarum delisei Bory  37-  Phasganon macropterum Rupr.  9.  Alaria dolichorachis KLiellm.  42.  Laminaria musaefolia De l a P y l .  10,  A. dolichorachis f. longipes Miyabe  45.  L. musaefolia j2> remotifolia De l a P y l .  21.  A. esculenta f. Stroemf.  49.  4. oblonga Kjellm.  22.  Fucus esculentus v a r . ^ minor Turner  52.  PI. paradiseum f. brevipes Miyabe et Nagai  23.  F . Teres Good, et Woodw.  53.  Orgyia pinnata (Gunn.) Ghobi  24.  Pleuropterum fasciculatum Yamada  62.  A. tenuifolia f. amplior Setch.  29-  A. fistulosa f. platyphylla Setch.  64.  A. valida K.iellm. et Setch.  30.  A. fragilis Saunders  65.  A. valida f. longipes Setch.  31.  A. f r a g i l i s f. bullata Saunders  66.  A. yezoensis Miyabe  34.  A. lanceolata Kjellm.  fasciculata  -228-  Figure 60  Figure 6l  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items