A CONTRIBUTION TO THE CLASSIFICATION AND PHYLOOENY OF THE MARTEN (SUBGENUS MARTES PINEL) OF NORTH AMERICA WITH SPECIAL REFERENCE TO THOSE OF BRITISH COLUMBIA. by EDWIN MOYER HAGMEIER A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS, IN THE DEPARTMENT OF ZOOLOGY. We accept this thesis as conforming to the standard required from candi-dates for the degree of .Master-of Arts. Members of the Department of THE UNIVERSITY OF BRITISH COLUMBIA April , 1952 I CONTENTS page CONTENTS .1 ILLUSTRATIONS V " 1 1 1 INTRODUCTION . 1 METHODS 2 PART I. Classification . ... ... 3 Results 3 Genus Maries Pinel 1792 . 4 Subgenus Maries Pinel 1792 6 Subspecies group 1., americana 10 Maries americana americana (Kerr 1792) 1G Maries americana brumalis (Bangs 1898) ..... 15 Maries americana airaia (Bangs 1897) ...... 17 Maries americana abieiicola (Preble 1902) .. 18 Maries americana aciuosa (Osgood 1900) 20 Maries americana kenaiensis ( E l l i o t 1903) .. 23 Maries americana abietinoides (Gray 1865 .... 24 Subspecies group 2., ^aurina 27 Maries americana caurina (Merriam 1890) ... 27 Maries americana origenes(Rhoads 1902) ... 30 Maries americana humboldtensis Grinnell and Dixon 1926 ...... 32 Maries americana sierrae Grinnell and Storer 1916 -. 34 i Maries americana vancouverensis Grinnell and Dixon 1926 ...... 35 II CONTENTS continued page Martes americana nesophila (Osgood 1901) 37 Conclusions 39 PART I I . Phylogeny 41 Discussion 41 1. Classification and palaeontologies! data 42 2. Palaeoecological data 45 Conclusions 53 SUMMARY 53 LITERATURE CITED 56 APPENDICES • ; 78 A. Definitions of measurements used 78 B. Provisional a r t i f i c i a l key to skulls of the subspecies of North American marten 79 ABSTRACT 84-1X I l l ILLUSTPATIONS Page F i g . 1. Original distribution of Eurasian, marten 6 F i g . 2. Original distribution of North American marten 9 Fig."3. Original distribution of marten in B r i t i s h Columbia ..... 10 F i g . 4. Distribution in time of the Mustelidae 43 F i g . 5. Marten distribution lying outside of regions of glaciation 49 F i g . 6. Number of marten trapped per 1000 square miles of forest 50 F i g . 7. Assumed post-Wisconsin marten migrations 51 F i g . 8. A hypothetical phylogeny of North American marten 53 Table I. Former interpretations of numbers of species and subspecies of North American marten ...... 7 - "8i- - jX2T» ABSTRACT S t a t i s t i c a l study of 418 marten skulls gives evidence for the subspecific status of brumalis, americana, actuosa, caurina and nesophila. Abietinoides i s probably separable from americana. Vancouverensis, caurina and origenes cannot however be separated. Geographic distribution of the latter three suggests that for the time being their status should not be changed. Only one species occurs in North America, according to the unpublished work of Wright. This i s divided into two groups, referable to the formerly conceived species americana and caurina. The genus Martes arose in the lower Pliocene and segregated into the subgenera of martens and fisher during the Pliocene. The americana group preceded caurina to North America, both arriving sometime before the Kansan. Between the Kansan and the Illinoian the americana group divided into two branches, and by the Wisconsin or earlier, origenes had separated from caurina. > During the Wisconsin ice advance marten found refuge in four regions: the lake states, the Rocky Mountains, the Coast Range and Alaska. Repopulation of North America with the ice retreat occurred from these refugia, during which time or later f i n a l subspeciation took place. A CONTRIBUTION TO THE CLASSIFICATION AND PHYLOGENY OF THE MARTEN (SUBGENUS MARTES PINEL) OF NORTH AMERICA "WITH SPECIAL REFERENCE TO THOSE OF BRITISH COLUMBIA INTRODUCTION The present study has been carried out for the purpose of increasing knowledge concerning (1) the classification and (2) the phylogeny of the Martens (subgenus Maries Pinel) of North America, with emphasis on those of B r i t i s h Columbia. Past studies which have been made on the subgenus, as far as i t occurs in North America have been few. Rhoads (1902) has published the only paper concerned solely with this group and is the source of most subsequent references to i t . Other authors of 'impor-tance are referred to throughout the paper following. Sincere appreciation for assistance in .this study i s due Dr. I. McT. Cowan under whom i t was made, for advice given and for help in obtaining the specimens studied; to Dr. W. A. Clemens for ad-vice and encouragement; and to Mr. Y. Edwards, Mr. W. Cottle and Mr. R. Webb who put specimens at my disposal. Thanks are also due the custodians of the various museums who have made specimens available to me. Dr. P. L. Wright has kindly made unpublished findings of his own available to me, for which I am indebted. I wish to thank my wife as well for continued assistance and encouragement. METHODS The present study has been divided into two parts, as des-cribed above: (i) classification and (II) phylogeny. Classification in this paper i s concerned mainly with skull characteristics. Four hundred and eighteen s k u l l s / from the collection of the University of B r i t i s h Columbia Museum, the B r i t i s h Columbia Pro-vincial Museum, the National Museum of Canada, the Royal Ontario Museum, the Museum of the University of Michigan, the Carnegie Museum and the Museum of Vertebrate Zoology, were studied, employing the s t a t i s t i c a l methods"presented by Simpson and Roe (1939). . r . Originally a series of 32 standardized measurements, ratios and sums were made on each skull from B r i t i s h Columbia. Parameters for these were calculated and compared. Those which did not show sig-nificant differences between samples or were possessed of a : high coeffi-cient of v a r i a b i l i t y . ^ r e discarded. Eventually six skull characteris-t i c s were found to be of use. These are used throughout this study, and include ( 1 ) condylobasal length (2) rostral width (3) upper tooth row ( 4 ) bulla length ( 5 ) ratio of canine width divided by bulla length and ( 6 ) the sum of the width plus the length of the upper molar. The precise definitions of these are given in the Appendix. The f i r s t and third measurements, because of their variability could be taken only on adult skulls. Immature specimens were separated from mature ones - 3 -using the methods described by Rhoads (1902) and Marshall (1951). Because the characters used in obtaining the ratio in measurement (5) and the sum in measurement (6) showed a small coefficient of varia-b i l i t y themselves, i t was assumed that di f f e r e n t i a l growth rates did not affect their use together. For each sample measured, the mean, standard deviation, coefficient of va r i a b i l i t y and their standard errors, were calculated and the samples compared s t a t i s t i c a l l y . A l l measure-ments are given in centimeters. The conclusions concerning phylogeny presented i n the second part of this paper were arrived at after (1) examination and comparison of the skulls mentioned above and (2) a careful review of the pertinent literature. A bibliography of about eight hundred t i t l e s concerned with the classification, phylogeny and ecology of the marten has been pre-pared during this study. It i s expected that this w i l l be deposited in the library of the University of B r i t i s h Columbia. PART A. CLASSIFICATION.:. • Results: _ •Samples of- skulls-(sometimes discouragingly small) referable to the following .supposed populations (subspecies) were examined: -nesophila, vancouverensis. caurina, origenes, sierrae, humboldtensis, actuosa-i Boria, abieticola, abietinoides, americana and brumalis. No specimens representative of supposed atrata or kenaiensis were observed. - 4 -GENUS MARIES PINEL 17.92 1758 Mustela 1798 1765 1775 1777 1792 Z it> a line Fouine Maries Mustella Maries, Linnaeus, v o l . 1, p. 45; type Mustela Maries Linnaeus Buffon and D'Aubenton, vol. 7, p. 309 Buffon and D'Aubenton, vol. 13, p. 161 Frisch, p. 11 Scopoli, p. 491 Pinel, v o l . 1, p. 55; type, Maries domestica Pinel 1800 Vivsrra Shaw, vol. 1, p. 1820 Martes, " Nilsson, vol. 1, p. 38 1829 Zibellina Jiaup, p. 31 1848 Charonia Gis t e l , p. 559 1865 Pekania Gray, p. 107 1865 Charronia Gray, p. 108 1865 Foina Gray, p. 108 1911 Martes Thomas, p. 139 1928 Lamprogale Ognev, p. 26 The genus Martes has been recognized under different names as an entity since Linnaeus' 10th edition. These authors formerly giving i t the name Mustela called weasels Puiorius. Mustela is now reserved for the weasels and Puiorius for the polecats. The authority for the name Martes is at present i n some doubt. Buffon's terminology has never been considered Linnaean and for thai reason illegitimate. Thomas (1911) and Brongersma (1941) likewise - 5 -invalidate Linnaeus' Mustela as the name, the former stating that "the ermine was clearly the Mustela of both Gesner and Linnaeus and should therefore be treated as the type species". Alton (1879) states that "several systematists ... have assigned the Linnaean t i t l e Mustela to the martens ... on the ground that the ...' name was only employed ... to mark sous-genres, and was not used binomially to indicate genera. The f i r s t definite separation was made three years later by Nilsson, who gave the generic name Martes to the present group...." Palmer (1904) hov/ever, gives credit f o r t h e name to Frisch (1775). Thomas and Mi l l e r (1905) reported that Frisch had not been sufficiently pure in his usage of binomials, and was thus invalidated. Hershkovitz (1948) discredited Frisch for the same reason. Pinel (1792) is the next authority to use the name Martes. There i s however, even here, some question as to Pinel's absolute purity. Thus we arrive at Nilsson (1820) again. If Nilsson should not pass then the name martes would be lost, for the next to employ the name used i t for a group of viverrids. (Wagler 1830). Until f i n a l judgment is passed by the International Commission, authority for the name must remain open to question. For the purpose of this paper, however, i t seems best to consider Pinel the authority for the name Martes as do most current students. Most systematists consider the genus to be comprised of three subgenera! Pekania Gray 1865, Tomictis Hall 1931, and Martes Pinel 1792. SUBGENUS MARTES PINEL 1792 1792 Martes Pinel, v o l . 1, p. 55. The subgenus Martes is constituted of the true martens. The c present study recognizes one species as existing i n North America. Schmidt (1943) l i s t s three species (foina, martes and zibellina) from Eurasia ( F i g . l ) . Of interest ; i n this respect i s Pojaomarev's (1946) successful hybridization of martes and zibellina. Thus probably between two and four species (in the modern sense) of marten exist in the world to-day. However, future taxonomic work w i l l possibly result in a lowering of this number. Mayr (1942) points out that M i l l e r in 1912 considered 76$ of his species monotypic. While Okland in 1937 li s t e d only 55$ monotypic. There appear to be far fewer monotypic species of mammals than has formerly 1been recognized and this doubtless holds in the case of the Eurasian marten, as i t has in those of America. In the past, from one to half a dozen species of marten have been considered to occupy North America (Table I) •. Fourteen subspecies have been described as occurring but never a l l sat any one time or by any one author. Rafinesque (1819) considered the New ¥/orld marten a distinct species as did DeKay (1842). Richardson (1829) considered i t only a variety of the European Mustela martes and Brandt (1855) a variety of the Asiatic Mustela zibellina Baird i n 1857 and 1859 named one species M. americana as occurring in North America, and this distinct from Old World forms. Fig. I follow page F i g . l.» Original distribution of Eurasian Marten. Not shown i s the subspecies of Zibellina inhabiting Sakhalin Island. From Stirton (1939) after Ognev (1928), and from Schmidt (1943). V *m M. FOINA. ....•.•» M. M A R T E S . M. ZIBELLINA. 7 That he was i n some doubt as to i t s exact status i s revealed by his statement (1859) that ^1 am ... inclined to the belief that we have two species, one representing the pine marten ... the other similar to the sable ... both probably distinct from the corresponding Old World species." Gray in 1865 distinguished between old and new world forms by the small size and peculiar shape of the upper molar i n the lat t e r . He separated the one species, americana, into three "varieties": abietinoides whose "habitat" was the ''Rocky Mountains," huro of "Fort Franklin" and leucopus whose habitat is not named. I have found i t d i f f i c u l t to relate Gray's varieties to the forms designated by more recent workers. In 1874 Ames l i s t e d but did not describe two American species. M. americana and M. martinus Allen in 1869 refused to accept the presence of an American marten distinct from the Eurasian. He believed "beyond reasonable doubt that ... a l l thus far described belong to a single circumpolar species, with possibly one or two ... continental races." He called , the American form Mustela martes Linnaeus. In 1876 Allen acknowledged the new world form as a distinct species on the basis of the lack of an "inner cusp on the second lower molar," and named i t Mustela americana. At the same time he pointed out the increase i n size of the skulls of marten taken from more northern latitudes. Coues (1877) and True (1884) named only one species, M. ameri-cana. Table I follow page Table I. Former interpretations of the numbers of species and subspecies of North American marten. The bracketed numbers represent the number of subspecies list e d under that particular species. In the two columns at the right are the total number of species and subspecies named by each author. Note that Anderson (1946) considered only Canadian forms. Richardson 1829 Brandt l85+2 Anderson 19^6 Mustela martes Mustela z i b e l l i n a var. americana Mustela americana Martes americana (3) Mustela martes Mustela americana. M. martlnus Mustela americana Mustela americana Mustela americana Mustela americana Mustela americana (2) Mustela americana (*+) Mustela americana (5) Martes americana (5) Martes americana (?) Martes americana (5) Martes americana (7) Martes americana (5) atrata (W) atrata atrata atrata atrata M. caurina M.caurina (2) M. caurina (2) M. caurina (2) M. caurina (3) MV atrata M. caurina (3) M. atrata M. caurina (6) M. caurina (3) M. atrata M. brumalis M. brumalis M. brumalis M. brumalis M. boria M. boria M. boria M. kenaiensis M. nesophila M. nesonhila M. nesophila 0 . Q. 0. CO t o CO 1 1 1 1 3 1 2 1 1 1 2 5 U 5 3 7 5 10 6 11 6 12 5 11 2 13 3 9 / 8 With E l l i o t in 1901 the new era of taxonomic splitting had begun, probably originating with Merriam's (1890) description of caurina. This trend was continued through Mi l l e r and Rehn (1901) to Rhoads (1902) the first and last monographer of the groups. Working on 129 American and several foreign skulls, Rhoads divided the new world marten into three species and seven subspecies. E l l i o t in 1905 li s t e d six species and eleven subspecies. In 1912. Mi l l e r on the basis of Thomas! (1911) findings, changed the generic name from Mustela to Martes. Miller's (1924) catalogue is to-day considered the standard in systematics. He1 l i s t s six species and twelve subspecies. Anthony in 1928, Allen in 1942 and Anderson in 1946 a l l made minor changes in the classification of the group, Allen l i s t i n g only two species. At present i t appears l i k e l y that only one species of marten exists in North America, as was suggested as early as 1946 by Hall. Dr. P. L. Wright (letter of July 9th 1951) informs me that s t a t i s t i c a l treat-ment of 250 skulls from Idaho and Montana shows complete intergradation between the two primary subspecies groups of American marten, thus indi-cating that they are probably conspecific. If these two major subdivis-i o n s - f a i l to show genetic isolation i t i s very l i k e l y that none exists in the remaining less differentiated groups. Wright's work, as well as other signs of intergradation noted i n this study, lead to the conclusion that in a l l probability only one species inhabits the contin-ent. The present study shows that a l l American marten f a l l into two main subspecies groups, namely the americana group (formerly Martes americana) and the caurina group (formerly M. caurina). The original distribution of marten kept "very close to the Canadian Faunal Region" (Seton 1929). That i t s distribution corresponds very closely to that of the narrow-leaved evergreen l i f e form of tree i s well known. (See Seton's 1929 map) Jurgenson (1939) reported that marten correlate well with the distribution of f i r forests of the European U. S. S. R., but that preferred even to this is a mixed deciduous-coniferous forest ("Piceatum composition"). Emmons (1840), Rhoads (1903), Allen (1904) and Morris (1948) however remark that marten are p a r t i a l to hardwood as well as coniferous for-ests. The distribution of the marten, as i t is assumed to have existed in North America prior to the advent of the white man, i s mapped in F i g . 2 . A l l faunal l i s t s known and available to me have been used in i t s compilation. The mappings of the distribution of the Boreal and West Coast coniferous floras have been used in delimiting much of the periphery of the range, however. Ea r l i e r maps have been prepared by Seton (1929) and Anderson (1934) for North America, and by Cory (1912) and Hamilton (1943) for the eastern half of the continent. F i g : 3 maps in analogous fashion the distribution limits as understood at present i n B r i t i s h Columbia. It is well known that the pres ent range of marten is far different from what i t was prior to the coming of the white man to the New World (See Allen 1942, Seton 1929). Huxley (1940) has mapped the present range of the Asiatic marten (after Ognev, 1928-30) as has de Vos (1951) for those of Ontario. Both show the populations of their respective regions to be broken into many (forty in Ontario) small isolated groups, none in contact with the other, and to possess a F i g . 2. (To follow page 9.) F i g . 2. Hypothetical distribution of marten in North America prior to the advent of white men; based on faunal l i s t s , state and continental range maps, and where these are inadequate, on the forest maps of Bowman (1911), Brown (1950), Halliday (1937), and , Shantz and Zon (1924). Circles mark type lo c a l i t i e s of the subspecies named as occurring in that area (from Anderson 1934b). The numbers represent the following subspecies: 1 Martes americana .atrata; 2 M. A. brumalis; 3 M. a. americana; 4 M. a. abieticola; 5 M. a. actuosa; 6 M.a. Kenaiensis; 7 M.a...nesophila 8 M.a. caurina; 9 M. A. vancouverensis; 10 M.a. abietin-oides; 11 M. a. origenes; 12 M.a. humboldtensis; 13 M.a. sierrae. Stippled areas represent supposed regions of inter-gradation. - 10 -southern limit far north of what i t originally was. The range maps presented with this study are meant to be only approximations of the original distribution of the animal concerned. The method of mapping areas of intergradation in these i s described i n the discussions of subspecies. SUBSPECIES GROUP 1., AMERICANA Nomenclature and classification; Formerly the species americana and i t s subspecies. Characteristics; Ratio of canine width at base - / -bullae length nearer to .90 than to 1.50 in males, .85 than .95 in females. Upper molar length width nearer to 1.20 cm. than to 1.35 cm. in males, .1.05 cm. than to 1.20 cm. in females. Distributiont A l l marten east of the Coast Range, excluding those of the Rocky Mountain Chain south of the 49th parallel and those of the Sierra Nevadas, are of this group. MARTES AMERICANA AMERICANA (KERR 1792) 1772 Mustela martes, Forster, p. 372 (after Linnaeus) 1776-80 Mustela zibellina Pallas, vol. 2, n. 57. 1792 Mustela zibellina americana Kerr p. 1802 Mustela americanus, Turton, v o l . 1, p. 60. 1819 Mustela vulpina Rafinesque, p. 82 1820 Mustela leucopus Kuhl, p. 74 1823 Mustela huro Cuvier, v o l . 29, p. 256 F i g . 3. (To follow page io.) F i g . 3. Hypothetical distribution of marten in Br i t i s h Columbia prior to the advent of white men; based largely on the faunal. maps of Edwards (1950) and Munro and Cowan (1947). Stippled areas represent regions of supposed intergradation. A C T U O S A A B I E T I N O I D E S C A U R I N A | V A N C O U V E R E N S I S I N E S O P H I L A I N T E R G R A D E S - 11 -1827 MustelaLeucotis G r i f f i t h , Smith and Pidgeon, vol.' 2, p. 297 and plate facing. 1827 Mustela s i n u e n s i s G r i f f i t h , Smith and Pidgeon, v o l . 2, p. 297. 1843 Martes leucopus, Gray, p. 63 1844 Mustela (Martes) huroSchintz, p. 337. 1844 Mustela (Marte3) leucopus, Schintz, p. 337. 1844 Mustela (Martes) vulpina, Schintz, p. 377. 1855 Mustela z i b e l l i n a var., americana Brandt p. 16. 1865 Martes americana vars. abietinoides, huro and leucopus Gray, p. 106. 1874 Mustela martinu3 Ames, p. 69. 1885 Mustela americana, True, p. 610 1912 Martes americana americana, Miller, p. 92. Type l o c a l i t y : "North America" (Turton 1802); "Eastern North America'* (Miller 1900 and most authors); "Upper Missouri ?" ( E l l i o t 1901); "The region occupied by the small pale martens of southern Canada and. the northern United States" (Rhoads 1902). Type specimen: None, to my knowledge, unless i t i s that of Mustela i martes described by Forster (1772) after.Linnaeus' description. Nomenclature: The authority for the name M. a. americana has generally been attributed to Turton (1806). Khoads (1902) however pointed out that Turton's 1802 edition just contained the name, and my examination of Turton's earlier edition confirms his conclusion. Allen (1895) stated that Kerr (1792), while not s t r i c t l y - 12 -Linnaean, was sufficiently so to make his terminology l e g i t i -mate. He stated that Turton copied many of his names, includ-ing "Mustela americana", from Kerr and that i t is only due to the scarcity of Kerr's book that these have been attributed to Turton, Although I have not examined Kerr, the authority for the name, on Allen's statement, must rest with him unti l other-wise i s shown. Classification: Although Baird, as has been pointed out earlier, was the f i r s t to accept Turton's designation, Gray (1865) was the f i r s t to base the separation on skull morphology - "the small size of the last tubercular molar" i n americana. Merriam in his description of caurina (189Q) list e d many characteristics peculiar to americana. However, credit goes to Rhoads (1902) for the f i r s t , and last, thorough analysis of the species. He found Coues', Gray's and Allen's separation of americana from martes, foina and zibellina on the basis of "the slight constriction with resulting rectangular shape" of the upper molar was f u l l y warranted. Further he affirmed Allen's state-ment that americana almost always lacked the inner cusp of the second lower molar. Rhoads found that marten from the Cascades and Pacific coast resembled foina however, far more than americana or Rocky Mountain forms, in the characteristics of the upper molar, f l a t cranium and broad zygomatic breadth. Great dissimilarity with zibellina was noted. On the other hand he found caurina - 13 -and foina similar in the small size of the bulla, and breadth and flatness of the brain case. In respect to sagittal crest, frontal constriction and teeth, he found closer relationship between^caurina and zibelli n a . He concluded that the caurina subspecies group i s connected "far more closely with Eurasian than with the American type of marten ..." and are isolated members of the "martes-zi b e l l i n a group".. Characteristics: Next to abietinoides, from which i t i s separated with d i f f i c u l t y , i t i s probably the smallest marten of the americana group. The two, however, are geographically isolated by inter-mediate populations of abieticola and actuosa. MALES M S D C V N-C L 7.982 .045 .223 .032 : 2.8 .4 24 R W 1.479 .007 .063 , .005 4.3 .4 68 U T R 2.930 .021 .101 .014 3.4 .5 23 8 L 1.677 .004 .035 .003 2.1 .2 67 CW/BL .847 .003 .032 .002 3.8 .3 67 MW L 1.215 .007 .062 .005 5.1 .4 67 FEMALES C L 7.252 .086 .258 .060 3.6 .8 9 R W 1.300 .010 .064 .007 4.9 .6 39 U T R 2.683 .036 .109 .025 4.1 .9 9 B L 1.564 .010 .068 .007 4.3 .5 39 CW/BL .805 .006 .038 .004 4.7 .5 39 MW L 1.077 .008 .056 .006 5.2 .6 39 - 14 -The symbols in the preceding table" represent the following charac- ] ,ters: G L 3 Condylobasal Length; R W = Rostral Width; U T'R = Upper i Tooth Row; B L " B u l l a Length; CW/BL « Canine Width «.,/ -Bulla Length; MW L ° Upper Moiar Width Length. These symbols are used i n a l l descriptive tables following. S t a t i s t i c a l difference of significance (to the .05 level throughout this paper) i s shown between the means of the samples of americana and brunalis in the ratio of canine to bulla length (males and females), condylobasal length (males only), rostral width (males only), upper tooth row (males only), bulla length (males only) and upper molar length plus width (males only). A similar significant difference between the means of abieticola and americana i s shown i n rostral width and upper molar width plus length (males only). Abietinoides shows a significant difference from americana only, in the ratio canine width - / - bulla length (males only). i . Distribution: Original range, as far as known, as mapped. Four specimens from Lake Mistassini show characteristics intermediate between americana and brumalis. Five specimens from Sioux Lookout D i s t r i c t show none of the characteristics of abieticola. The boundaries of the range americana mapped in F i g . 2, since they do not disagree with the facts presented in this study, are taken from Cory (1912). Formerly i t probably ranged as far south as northern Virginia (Audubon and Bachman 1856). Its southern limit- i s now probably Algonquin Park (Downing 1948) or possibly the Berk-shire Mountains of Massachusetts (Hamilton 1943) De Vos (1951) has mapped it s present range in Ontario. > - 15 -Specimens examined: Gaspe Peninsula, five (N M C); A b i t i b i D i s t r i c t Quebec, three (N M C); Kapuskasing, Ont:-. one (N M C)j Presque Isle Co., Mich, one (U.M)j Sioux Lookout Distr. Ont. five (ROM Z); Kenora Distr., Ont. two (R 0 1 Z); Algoma Distr., Ont. 12 (R 0 M Z)j Sudbury Distr., Ont. 59 (R 0 M Z); Nipissing Distr., Ont. five (R 0 M Z); Thunder Bay Distr., Ont. seven (R 0 M Z); Cochrane Distr., Ont. four (R 0 M Z); Renfrew Distr., Ont. one (ROM Z); Lake Mistassini, Que. four ~(N M C)^r Muskoka Twp., Ont. one (R 0 1 Z); In the above, and in;,all future discussions, the following abbreviations are used in naming the collections studied: N M C National Museum of Canada; U M University of Michigan Museum; R 0 M Z Royal Ontario Museum of Zoology; C M Carnegie Museum; U B G University of B r i t i s h Columbia Museum; B C P M B r i t i s h Columbia Provincial Museum; M v Z Museum of Vertebrate Zoology. MARTES AMERICANA BRUMALIS (BANGS.1898) 1898 Mustela brumalis Bangs, p. 502 r r 1901 Mustela atrata brumalis, E l l i o t , p. 336. 1902 Mustela americana brumalis, Rhoads, p. 448 1909 Mustela caurina brumalis, Hantzch, vol. 9, p.254. 1912 Martes brumalis, Miller, p. 93. 1934 Martes americana brumalis, Anderson, p. 95 Type localitys "Okkak, Labrador" (Bangs 1898)j "Okak, Labrador" (Miller 1900) . Type specimen: Museum of Comparative Zoology, Bangs' Collection No. 7417, Cambridge, Mass. (Bangs 1898, Anderson 1946). . -- 16 -Characteristics: "when compared with skulls of true americana ... show very striking differences; their large size, short wide rostrums and enormous teeth at once distinguish them" (Bangs 1898) Rhoads (1902) considered the skulls of brumalis "surpris-ingly identical in size and proportion to those ofnactuosa" and the present study bears out his conclusion. On the basis of the small sample studied i t is the larg< North American marten, being most similar to actuosa. MALES M S D C V N C L 8.643 .118 ^204 2.4 3 R W 1.619 .017 .045 2.8 7 U T R 3.190 .026 .046 1.4 3 B L 1.749 .019 .051 2.9 7 CW/BL .906 .012 .033 3.6 7 MW L 1.363 .018 .048 3.5 7 FEMALES. C L 1.550 .149 .212 2.8 2 R W 1.475 .074 .106 7.2 2 U T R 2.795 .016 - .023 .8 2 B L 1.560 .060 .085 5.4 2 CW/BL .900 .009 .014 1.6 2 'MW L 1.150 .029 .042 3.7 2 - 17 -S t a t i s t i c a l comparison with americana has been made under that subspecies discussion. Compared with actuosa (the only one similar to i t ) s t a t i s t i c a l significance of difference between means is shown in upper molar length plus width (males only). On the basis of the samples at hand no other differences have been detected. Distributions Labrador probably as mapped, north to the tree limit, intergrading with americana in the region about Lake Mistassini at least. Newsom (1937) records the presence of marten on Anticosti Island, but that they d i f f e r from the mainland form in having the "fore-paws and end of t a i l ... tipped with white hair." Specimens examined: Fort Chimo, Ungava four (N M C); Grand F a l l s , Hamilton R., Lab. two (C M); v i c i n i t y Northwest R., Lab. three (CM) MARTES AMERICANA ATRATA (BANGS 1897) 1897 Mustela atrata Bangs, p. 162. 1912 Martes at r a t a . M i l l e r , p. 93 1942 Martes americana atrata, Allen, p. 166. -Type lo c a l i t y : "Bay St. George, Newfoundland" (Bangs 1897) Type specimen: Museum of Comparative Zoology, Bangs' Collection No. 5752, Cambridge, Mass. (Anderson 1946). No specimens of atrata have been examined i n this study. Bangs (1897) states however: "Size about that of M. americana (probably somewhat larger)} colour very different ... skull slightly different ... - 18 -about the same size or larger ... ; rostrum narrow, audita! bullae much larger and deeper ... and with a more marked 'bottle-nose' pro-jection; dentition rather weaker throughout, with greater spaces between premolar teeth." Rhoads (1902) stated that Bangs considered this a species on geographic rather than morphologic grounds, and that i t is cranialiy much closer to americana than to brumalis. Most authors consider atrata to be a f u l l species however. Gilpin (1860) remarked that pelages of marten from Newfound-land were much more similar to those of Labrador (brumalis) than to those of Nova Scotia (americana)• Since Vancouver and Queen Charlotte Islands both bear separate subspecies of marten i t is quite possible that examination of Newfoundland specimens w i l l show them to be subspecifically d i f f e r -ent from mainland forms as well. On the basis of Bangs' description however, atrata cannot be given specific status and on the basis of i t s geographic position and in the interests of parsimony i t must be considered a subspecies of the americana subspecies group unt i l shown to be otherwise. Distributions Newfoundland. Specimens examined: None. MARTES AMERICANA ABIETICOLA. (PREBLE, 1902) 1902 Mustela americana abieticola Preble, p. 68. 1912 Martes americana abieticola, M i l l e r , p. 92 - 19 -Type l o c a l i t y : "Cumberland House, Saskatchewan" (Preble 1902) Type specimen: U. S. National Museum, No. 19256/34962 (Poole and Schantz , 1942) Characteristics: Described by Preble (1902) as "much longer than Mustela americanus ... the sagittal crest being more highly developed ... dentition much heavier except last upper molar which i s usually about the same size.... Approaches Mustela a. actuosa but though smaller has heavier dentition." The several specimens examined during this study show characteristics intermediate between those of americana and actuosa. MALES M S D C V N C L 8.27 ;(;;v....-...i V.;..... 1 R W 1.625 .015 .022 .011 1.4 .6 2 U T R 3.05 1 B L 1.690 .069 .098 .049 5.8 2.9 2 CW/BL .962 1 M L 1.390 .019 .028 .014 2.0 1.0 2 S t a t i s t i c a l comparison with americana has been made under the description of that subspecies. Significant differences in the comparison of means of actuosa with abieticola were found in upper molar length plus width (males only) Specimens from Fort Vermilion and Wood Buffalo Park, Alberta and Thelon Game Sanctuary, N. W. T. appear much nearer to actuosa than abieticola. i t i s quite probable that abieticola should be considered - 20 -an intergrading form separating actuosa from americana. Because of the smallness of the sample studied, i t i s best to maintain i t s f u l l subspecific standing un t i l proven otherwise. Distribution: Western Northern Ontario (?), Manitoba, Saskatchewan, and possibly extreme eastern Alberta and Mackenzie d i s t r i c t . Specimens examined: Herchmer, Man. two, (N.M.C). MARTES AMERICANA ACTUOSA (OSGOOD 1900) 1900 Mustela americana actuosa Osgood, p. 43 1901 Mustela atrata actuosa, E l l i o t , p . 336. 1905 Mustela americana actuosa, E l l i o t , p. 139. 1908 Mustela americana actuosa, Preble , pp. 236-237 1912 Martes americana actuosa, Miller, p. 93. 1912 Martes boria Miller, p. 93 1937 Martes actuosa, Anderson, p. 1942 Martes americana boria,. Allen, p. 167 Type l o c a l i t y : "Fort Yukon, Alaska" (Osgood 1900) Type specimen: U. S. National Museum No. 6043 (Poole and Schantz 1942). Classification: Three specimens from Mackenzie River d i s t r i c t , the approxi-mate type loc a l i t y of boria, show no significant difference from those from Alaska, and constitute justification for accepting Preble's (1908) conclusion that boria and actuosa represent different colour phases of the same subspecies. - 21 -Characteristics: "Similar to M. brumali3 but larger; ... skull somewhat larger, longer; dentition relatively much weaker; last upper molar decidedly the largest of the subspecies of Mustela americana. M. brumalis i s also large but does not equal actuosa ... and has heavier dentition." (Osgood 1900). Rhoads (1902) found that he could not distinguish brumalis from actuosa according to Osgood* s diagnosis and the present study bears him out. Brumalis, further, on the basis of the small sample studied, appears the larger of the,two. Further comparison i s given under the description of that subspecies MALES M S D C V N C L 8.325 .076 .255 .054 3.1 .7 11 R W 1.626 .014 ' .060 .010 3.7 .6 17 U T R 3.041 .027 .104 .019 3.4 .6 14 B L 1.753 .014 .056 .009 3.2 .6 16 CW/BL .821 .012 .048 .008 5.2 .9 i s ISM L 1,249 .015 .064 .010 5.1 .9 17 FEMALES C L 7.677 .074 .166 .052 2.2 .7 5 R W 1.434 .016 .043 .011 3.0 .8 7 U T R 2.780 .020 .045 .014 1.6 .5 5 B L 1.619 .020 .053 .014 3.3, .9 7 CW/BL .869 .009 .026 .006 3.0 .8 7 MW L 1.136 .013 .037 .009 3.3 .9 7 - 22 -S t a t i s t i c a l comparison with subspecies whose ranges l i e to the east of actuosa has been made. Compared with abietinoides to the south, significance between means was found in condylobasal length (males).,, rostral width (males and females), bullae length (males and females), upper molar width plus length (females), and upper tooth row (females) . Twenty-seven specimens from north central B r i t i s h Columbia, show characteristics intermediate between actuosa and abietinoides. Even though these-show an average coefficient of v a r i a b i l i t y no higher than the latter, and the former do (3.7 compared to 4.5 and 3.3), they are probably best classed as intergrades. One skin from Bulkely House, Br i t i s h Columbia (B C P M) was quite different from five actuosa skins from Yukon (M V Z) and from six abietinoides skins (U B C) from south eastern B r i t i s h Columbia. The Bulkely House specimen was grayer, and with a whiter throat patch than in those of either actuosa or abieti-noides, but in other characteristics showed greater similarity to actuosa. Twenty skulls from Fort Nelson B. C. are, however, typical actuosa. In a similar fashion, eight skulls from Fort Vermilion, Alta. and Wood Buffalo Park, Alta . show characteristics intermediate between actuosa and abietinoides. Distribution: A l l of timbered Alaska, excepting the Kenai Peninsula (sup-posedly), Yukon and the Northwest Territories, apparently intergrading with abieticola at about the 110th parallel longitude, and with abieti -noides in the area ranging from about the 60th parallel latitude south to about the region of the continental divide. Absent from the Rocky - 23 -Mountains south of 55 and from the Coast Ranges. Specimens examined: Fort Vermillion, Alta. five (N M C); Wood Buffalo Park, Alta . three (N M C); Thelon Game Sanctuary, N W T. one (N.M C); Fort Nelson B.C. twenty (U M); lower Iskut R., B.C. one B C P M); Brit i s h Columbia four ( N M C, D B C)j Fort Graham, B.C. seven (NMC)j Takla landing, B.C. three ( B C P M)-; Driftwood R., B.C. three (B C -P M); Omineca R., B.C. two (U B C); Bulkely House, B.C. two (B C P M); Charley Creek, Yukon R., Alaska five (M V Z); White- R., Y.T. two (R 0 M Z); Mackenzie Delta, N.W.T. one (N M C); Kanto Rashke ?, Alaska two (M V Z); Chicton Mt., Alaska one (M V Z); Swanson Creek, Alaska three (M.V Z); 30 miles up ' Willow R., N W.T. two (N.M.C);, Nekukah, Alaska one (M V Z); Alaska six (M V Z , u M); Mackenzie Distr. N.W.T. one (N M C) MARTES AMERICANA KENAIENSIS (ELLIOT 1903) 1903 Mustela americana kenaiensis E l l i o t , p. 151 1905 Mustela kenaiensis, E l l i o t , p. 421. 1912 Martes americana kenaiensis, Miller, p. 93 Type l o c a l i t y : "Kenai Peninsula, Alaska" E l l i o t (1903) Type specimen:. Chicago Museum of Natural History No. 9847 (Sanborn 1947). Classification: No specimens of kenaiensis have been examined during this study. E l l i o t (1903) describes i t as "smaller than M. a. actuosa, with longer t a i l , shorter feet, and general color darker; audital - 24 -bullae large, ridged, rectangular, longitudinally produced. Molars heavy; sagittal crest prominent; no gular or abdominal patches ... Skull smaller than those of either actuosa or brumalis." E l l i o t ' s measurements are: condylobasallength 7.95 cm.; upper tooth row 2.5; length of bulla 1.7. Osgood (1901) considered skulls of kenaiensis far more similar to M.a. americana than to actuosa. E l l i o t ' s measurements do not allow accurate comparison with those of actuosa. From the description given of the large bulla however, i t is l i k e l y that kenaiensis represents the americana subspecies group. On the basis of E l l i o t ' s statement that the animal lacks gular or abdominal patches, i t i s probably best to keep i t as a distinct subspecies for the time being. Distribution: Kenai Peninsula, Alaska. Specimens examined: None. MARTES AMERICANA ABIETINOIDES GRAY. 1865 1865 Martes americana variety 1, abietinoides Gray, p. i06. 1902 Mustela americana abietinoides, Rhoads, p. 451. 1912. Martes americana abietinoides. M i l l e r , p. 93. Type l o c a l i t y : "Rocky Mountains" (Gray 1865); "Edge of the humid west-ern slope of the Rocky Mountains between Kicking Horse Pass and the Columbia River" (Rhoads 1902). 25 Type specimen: None, to my knowledge. Rhoads pointed out that Gray's abietinoides included the range of origenes and on the basis of Richardson's (1829) descriptions, that a l l of Gray's specimens were true abietinoides; hence Rhoads change of type l o c a l i t y . Characteristics: "Black-brown; ears pale; head grizzled with white hairs, more or less gray; throat yellow or yellow spotted; throat spot large or broken up into small spots; the head sometimes with only a few gray hairs, and the throat with only a few distinct small spots" (Gray 1865) Rhoads (1902) described i t as possessed of a skull specifi-cally identical with actuosa, and only slightly smaller i n size, being intermediate in this respect between americana and actuosa, teeth bullae and skull proportions precisely as in americana. The present study disagrees with Rhoads' description in that i t finds abietinoides the smallest marten of the americana group, but agrees with his statement about the general similarity of proportion and size between abietenoides and M. a. americana. It i s most like americana from which i t is d i f f i c u l t or impossible to separate on the basis of skull characteristics. MALES H [ S D C V N C L 7.858 .035 .206 .024 2.6 .3 34 R W 1.514 .007 .057, .005 3.8 .3 58 U T R 2.869 .038 .223 .027 7.8 .9 34 B L 1.648 .007 .059 .005 3.6 .3 60 CW/BL .882 .005 .040 .003 4.5 .4 59 MW L 1.195 .006 •050 .004 4.2 .4 60 26 -FEMALES M S D C V N C L 7.094 .091 .491 .064 6.9 .9 29 R W 1.342 .009 .060 .006 4.5 .5 42 U T R 2.555 .020 .108 .014 4.2 .6 29 B L 1.537 .007 .050 .005 3.3 .4 42 CW/BL .835 .006 .040 .004 4.8 .5 42 MW L 1.057 .006 .045 .004 4.3 .5 42 S t a t i s t i c a l comparison with actuosa and americana has been made under the descriptions of these two subspecies. Thirteen speci-mens taken from the periphery of abietinoides range (Wisteria P.O., B.C., Tatuk Lake, B.C. and Tahtsa Lake, B.C.) are typical, of the subspecies; Distribution: B r i t i s h Columbia south of the continental divide, intergrading with actuosa to the north; east to the eastern slopes of the Rocky Mountains, intergrading with actuosa and possibly abieticola; west to the eastern slopes of the coast range,, probably intergrading with caurina-and south into northern Idaho and Montana, intergrading with origenes. Specimens examined: Tahtsa L., B.C. one (Collection of Robert Webb); Tatuk L., B.C. one (N.M C); Wisteria P.O., B.C. eleven (B C P M); Okanagan L., B.C. two (UB C); Jasper, Alta. one (N M C); Yahk, B.C. two (NM0)| Eagle R., B.C. eighteen (U B C); Akalkolex R., B.C. four (B C P M); Banff Park, Alta. sixteen (U B C); Rocky Mts. west of " Banff eight (U B C); Akamina R., B.C. fourteen (U B C); Blue R., B.C. seven (U B C)j Quesnel L., B. C. one (U B C); L i t t l e Prairie, B.C. two ( B C P I ) i Golden, B.C. one (U.B.C); Fatigue Creek, Alta . one - 27 -(U B C); Simpson Pass, Alta. twenty-one (U B C); Lempriere, B.C. one (U B C); Murtle L., B.C. one (U B C); Cranbrook, B.C. one (R 0 M Z); Morrissey, B.C. one (N M C); Nelson, B.C. one (N M C); Jasper. Park, Alta. two (U B C); Vermillion Pass, B.C. two (U B C). SUBSPECIES GROUP II, CAURINA. Nomenclature and classification; Formerly the species Martes caurina and' i t s subspecies. Characteristics; Ratio of canine width at base'--/'-bulla length nearer to 1.50 cm. than .90 cm. in males, .95 than .85 in females. Upper molar length plus-width nearer to 1.35 cm. than to 1.20-cm. in males, 1.20 than 1.05 in females. Distribution; The coast ranges and Sierra Nevada3 together with the Rocky mountains south of the 49th par a l l e l ; Vancouver and Queen Charlotte Islands and elsewhere as described below. MARTES AMERICANA CAURINA . (MERRIAM..1890). 1890 Mustela caurina Merriam, p. 27 1901, Mustela americana caurina. Osgood, p. 70. 1901 Mustela atrata caurina. E l l i o t , p. 335 1912 Martes caurina caurina, Miller, p. 93. 1951 Martes americana caurina, Wright, unpublished MS. Type l o c a l i t y : "Chehalis County, Washington (coast near Gray's harbor)" (Merriam 1890). Type specimen: U. S. National Museum Collection No. 186,450 (Poole and - 28 -Schantz 1942). Characteristics: "In external appearance ... differs l i t t l e from M. americana ... except that the irregular markings of the throat and under surface generally are orange-red instead of whitish or yellowish.... The skull ... differs ... in the following particu-la r s : the rostral portion i s broader and shorter; the audital bullae are shorter and less inflated; the frontals are broader both interorbitally and post-orbitally; the shelf of the palate i s less produced behind the plane of the last molar; the f i r s t upper pre-molar is smaller and more crowded; the upper molars are larger; the upper sectorial in addition to i t s larger size, has the inner lobe very much larger and longer, projecting anteriorly beyond the plane of the anterior lobe, the reverse being the case in M. americana; the last upper molar i s not only larger, but has a much broader saddle; the transverse diameter of the tooth i s about one-third greater than in M. -americana, and the antero-posterior diameter of the inner lobe is both relatively and absolutely much greater; the distance between the outer alveoli of the upper canines equals the greatest length of the audital bullae instead of being much less; the transverse diam-eter of the upper molar i s greater instead Of less than the length of the upper sectorial ... the underjaw i s heavier." (Merriam 189G). The present study finds that caurina, as well as a l l the other subspecies of the group, are in general well characterized by the above description. - 29 -MALES M S D C V N, C L 7.863 .066 .230 .046 2.9 .6 12 R W 1.639 .016 .072 .011 4.4 - .7 19 U T R 2.920 .028 .099 .020 4.4 .9 12 B L 1.471 .018 .080 .012 5.4 .9 19 CW/teL 1.041 .016 .070 . .011 6.7 1.1 19 MW" L 1.337 .011 .052 .008 3.9 .6 19 FEMALES C L 7.251 .071 .227 .050 3.1 .1 10 R W ' 1.479 .021 .082 .015 5.5 1.0 14 U T R 2.670 .026 .083 .018 3.1 .7 10 B L 1.427 .015 .059 .on 4.1 .8 14 CW/BL .981 .017 .063 .012 6.4 1.3 13 MW L 1.227 .019 .072 .013 5.9 1.1 14 One specimen from Yakutat, Alaska shows greater similarity to actuosa than to caurina in bulla length and the ratio of canine width to bulla length. Greater similarity to caurina is present i n rostral width and upper molar length plus width. Another specimen from Hole-in-the-wall, B.C. shows characteristics intermediate between those of caurina and actuosa. Distribution: The coast range of North America; exact northern and southern limits unknown, but supposedly as far north as the base of the Alaska - 30 pan-handle (Anderson 1946), and as far south as southern Oregon. (Grin-n e l l , Dixon and Linsdale 1937). Specimens examined: Hole-in-the-wall, B. C. one (U B C) j Yakutat, Alaska one (M V Z); Anahim L., B.C. six (M V Z~); Manning Park, B_.C twelve (U B C); Rainbow Mts, Mt. B r i l l i a n t , B.C. one (N M C)| Bella Coola, B.C. one (B C P M); Stuie, Cariboo Mt., B.C. one (N M C); Atnarko, B.C. one (B C P M); Lillooet distr., B.C. five (N M C, B C P M)j One-eye L., B.C. one (M V Z); Mason Co., Wash, one (U M); Chilliwack Valley, B.C. one (N M C); Kimequit, B.C. one (N M C)j Taliho, B.C. one (N M C); Kleena Kleena, B.C. one (M V Z ) ; coast of northern B.C. one (U B C); Kynoch Inlet, B.C. one (U B C). MARTES AMERICANA ORIGINES . (RHOADS 1902) 1902 Mustele caurina origenes Rhoads, p. 458 1 9- L 2 Martes caurina origenes, Miller, p. 93 1951 Martes americana origenes, Wright, unpublished MS. Type l o c a l i t y : " 'Marvine Mountain' (Garfield County?) Colorado" (Rhoads • .. 1902) Type specimen: U.S. National Museum Collection No. 112, 170 (Poole and Schantz 1942) Characteristics: "Strikingly different from a l l other American martens except caurina in the absence of the light coloured cheek patches. Head relatively darker ... than in caurina^ ears lacking a white bor-der.... Viewed from above shows an even greater uniformity of colour than does caurina" (Rhoads 1902). - 31 MALES M S D C V N G L 8.017 .074 .149 .052 1.9 .7 , 4 R W 1.643 .026 .076 .019 4.6 1.2 8 U T R 3,000 .036 .072 .025 2.4 .8 4 B L 1.531 .020 .057 .014 3.7 .9 8 CW/BL 1.015 .006 .017 .004 1.7 .4 8 Mff L 1.426 • - .024 .070 .017 4.9 1.2 8 FEMALES,. C L 7.260 .084 .147 .060 2.0 .8 3 R W 1.428 .020 .061 .014 4.3 1.0 9 U T R 2.636 .011 .020 .008 •8 .3 3 B L 1.418 .014 .043 .010 3.0 .7 . 9 GW/BL .930 .009 .026 .006 2.8 .7 8 MW L 1.231, .026 .076 .019 6.2 1.5 8 The present study shows no s t a t i s t i c a l l y significant d i f f e r -ences between the skulls of origenes and caurina; i t appears that Rhoads' determination of the subspecies was made wholly on pelage characteris-t i c s . Three skins of caurina (M V Z, U B C) and six of origenes (M V Z) have been examined and Rhoads' distinction between the two on the basis of these appears to, be only doubtfully valid. Because origenes i s completely isolated geographically, and because of the supposed d i f f e r -ences in coloration separating the two, i t seems best to maintain the - 3 2 subspecies u n t i l a larger sample of skulls and skins can be examined. Distribution:"" The Rocky Mountains of the United States, formerly south as far as Taos, New Mexico (Seton 1929). According to Wright (letter of July 9, 1951) i t extends as far north as Ravalli County, Montana> and central Idaho, where i t intergrades with abietinoides (of the americana subspecies group), marten from Glacier National Park and extreme northern Idaho being typical abietinoides. Dalquest (1948) identifies the marten of far eastern Washing-ton as origenes; Rust (1946) and Davis (1939) c a l l those of western Idaho caurina. In view of Wright's findings however, concerning the northern l i m i t of origenes, and the fact that a l l of southwestern 'British Columbian marten are abietinoides, i t is more likely that those of eastern Washington and north western Idaho are abietinoides, or abietinoides-origenes intergrades. M.a. origenes appears to be geographically isolated from caurina being separated from i t by the interior dry belt of western North America. " Specimens examined": Idaho eleven (M V Z, U M); Oregon one (M V Z ) ; Wyoming two (U M); Montana three (li Y Z ) . MARTES AMERICANA HUMBOLDTENSIS GRINNML AND DIXON 1926 1926 Martes caurina humboldtensis Grinnell and Dixon, p. 411. 1951 Martes americana humboldtensis, Wright, unpublished MS. - 33 -Type l o c a l i t y : "Ridge about five miles north east of Cuddeback =Carlotta , Humboldt County, California". (GrinneH and Dixon 1926) Type specimen: Museum of Vertebrate Zoology, Collection No. 19158, Berkeley (Grinnell and Dixon 1926) Characteristics: Has been described by GrinneH and Dixon as follows: externally "resembles Martes caurina caurina; decidedly darker, of richer golden brown tone than M. c. sierrae.... Skull smaller than i n either ... with rostrum slender, dis t i n c t l y constricted behind roots of canines; frontal region and postorbital constric-tion narrow; teeth small; molariform series not so crowded as in sierrae so as to skew the premolars; last upper molar with heel greatly expanded anteroposterior^, as much so, relative to trans-verse diameter... as in. sierrae, and hence more than in caurina" Three skulls of humboldtensis that arrived too late for f u l l consideration in this study are considerably smaller than the skulls of any other of the caurina subspecies group. No s t a t i s t i -cal difference between the means of the condylobasal lengths was found, however. Until more specimens are made available., and u n t i l the ones available are more carefully analyzed, i t i s best to maintain it: as a distinct subspecies. Distribution: The Cascade Range of California, south of about 38° l a t i -tude. According to GrinneEL, Dixon and Linsdale (1937) i t intergrades - 34 -with caurina "to the north of California". Bailey (1936) l i s t s Oregon martens as caurina however; i f the two represent distinct subspecies, i t must be admitted that the region separating them i s .unknown. Specimens examined: Humboldt Co., C a l i f , two (M V Z); Del Norte Co. Ca l i f , one (M v Z). MARTES ..AMERICANA.,. SIERRAE GRINNELL.,AND STORER...1916.. 1916 Martes caurina sierrae Grinnell and Storer, p. 2. 1926 Martes caurina nobilis Hall, p. 127 1951 Martes americana sierrae, Wright, unpublished MS. Type l o c a l i t y : "Head of Lye11 Canon, 9800 feet altitude, Tosemite National Park, California" (GrinneH and Storer 1916) Type specimen: Museum of Vertebrate Zoology No. 22112, Berkeley, California Grinnell and Storer 1916). Hall in 1936 placed nobilis, a Pleistocene subspecies described by him in 1926, in the synonomy of sierrae (p. 87). ( * extinct) Characteristics: Is described by Grinnell and Storer as "similar to Martes caurina caurina, but general coloration paler both above and below sides of face decidedly pale orange of chest very extensive.... Cranium with rostrum short as i n caurina but extremely narrow} whole cranium narrower and brain-case relatively higherj sagittal crest very weak} auditory bullae even smaller than in caurina" -35 -Five skulls of sierrae arrived too late for f u l l considera-tion in the present study. Qualitative examination shows them to be larger than those of humboldtensis, (but not significantly so), and to l i e within the range exhibited by caurina and origenes.One skin (M V Z) shows a summer pelage far paler, and acquired earlier, than do three of humboldtensis (M V Z). r As in the case of humboldtensis the subspecies must be allowed to stand. Distribution: The Sierra Nevada Range of California and extreme western Nevada. Specimens examined: Kern Co. C a l i f , one (M V Z); Emerald Co. C a l i f , one \ (M V Z)| Fresno Co. C a l i f , two (M V Z) j Madera Co. C a l i f . one (M V Z). MARTES AMERICANA VANCOUyERENSIS,, GRINNELL.,AND .DIXON 1926. 1926 Martes.caurina jvancouverensis Grinnell and , Dixon p. 414. 1951 Martes^americana Vancouver ens i s , Wright, unpublished ' MS. Type l o c a l i t y : "Golden Eagle Mine, 20 miles south of Alberni, Vancouver Island,,British Columbia^ (G^ 1926). Type specimen: Museum of Vertebrate Zoology, No. 12, 474, Berkeley (Grinnell and Dixon, 1926).. Characteristics: Described originally as "externally l i k e Martes caurina caurina but ... rostrum shorter, frontal region and indeed whole cranium - 36 -broader; molariform teeth more crowded and more skewed out of alignment. Last upper molar decidedly smaller, and i t s heel far less expanded anteroposteriorly. Skull in.general proportions more like that of Mustela nesophila Osgood but smaller, and with weaker dentition; last upper molar much smaller and heel even less expanded...." (Grinnell and Dixon 1926). MALES it [ S D C V N C L 7.968 .061 .154 .036 1.9 .5 9 R ¥ - 1.654 .020 .064 .014 3.9 .9 10 U T R 2.902 .020 <> .060 .014 2.1 .5 9 B L 1.519 .015 .050 .011 3.3 .7 10 CW/BL 1.029 .012 .040 .008 3.9 .9 10 MW L 1.329 .008 .028 .006 2.1 .5 10 - FEMALES C L 7.225 .038 .094 .027 1.3 .4 6 R W 1.468 .017 .045 .012 3.1 .8 7 U T R 2.605 .015 .037 .010 1.4 .4 6 B L 1.441 .014 .038 .010 2.6 .7 7 CW/BL .964 .015 .040 .010 4.1 1.1 7 MW L 1.127 .018 .048 .012 4.3 1.1 7 Vancouveren3is i s not, on the basis of the above samples, s t a t i s t i c a l l y separable from caurina. On a purely geographic basis - 37 -at least, and un t i l more skins of marten from Vancouver Island can be compared with those from the mainland, no change in i t s status seems to be warranted. Comparison with nesophila i s made under the description of that subspecies. Description: Vancouver Island, excluding a narrow strip supposedly along the.eastern and. southern shores. Specimens examined: Vancouver Island seventeen (B C P M, N M C, U B C)• MARTES AMERICANA,NESOPHILA. (OSGOOD 1901). 1901" Mustela nesophila Osgood, p. 33. 1912 Martes nesophila, M i l l e r , p.>94 1926 Martes caurina nesophila, Grinnell and Dixon, p. 417. 1951 Martes americana nesophila,Wright, unpublished MS.. Type l o c a l i t y : "Massett, Graham Island, Queen Charlotte Islands, B r i t i s h Columbia1.* (Osgood 1901). Type specimen: U. S. National Museum, Biological Survey Collection, No. 78066 (Poole and Schantz 1942).. Characteristics: "Similar to Mustela caurina but larger; rostrum shorter and heavier; dentition heavier; premolars larger and more crowded. Last upper molar similar to that of Mustela americana-, internal length being more nearly equal to external length than in caurina.... Especially separable by thick rostrum" (Osgood 1901). - 38 -MALES M - S D C V N G L 8.140 .084 .147 .060 1.8 .7 3 R W 1.863" .017 .030 .012 1.6 .7 3 U I R 2.983 .043 .075 .030 2.5 1.0 3 B L 1.496 .023 .040 .016 2.7 1.1 3 GW/BL 1.163 .025 .044 .017 3.8 1.5 3 MW L 1.353 .005 .010 .004 .7 .3 3 FEMALES C L 7.44 1 R W 1.60 1 U T R 2.70 1 B L 1.40 1 CW/BL 1.092 1 MW L 1.26 1 The above sample i s s t a t i s t i c a l l y separable from Vancouver-ens is on the basis of rostral width (males), canine width - / -bulla length (males and females) and upper molar length plus width (females). Similarly i t i s separable from caurina on the basis of rostral width (males) and canine width - / - bulla length (males). Swarth (1911 and 1912) after examining specimens from Admiralty Island and Kuiu Island, Alaska concluded that they might represent a distinct species. I have examined Swarth*s material however, and i t i s , for the characters studied, s t a t i s t i c a l l y • - 39 -comparable to nesophila and must for the present be considered representative of that .subspecies. Distribution: Known beyond doubt only on Graham and Moresby Islands, Q.C.I.» Admiralty Island and Kuiu Island, Alaska; mapped by Seton . as occurring on a l l of the Queen Charlotte Islands and a l l of the Alexander Archipeligo; i t is not however li s t e d as present on the coastal.islands between 51 and 54 degrees by McCabe. and Cowan (1945). Specimens examined: Graham Island, Q.C.I, four (B C P M, U"B C); Admiralty Island five (M V Z); Kuiu Island one (M V Z). Conclusions: On the basis of the s t a t i s t i c a l separability (to the .05 l e v e l or better) of the samples just described, there i s probably sound evidence for the existence of the following subspecies: americana, actuosa, abietinoides, caurina, nesophila and brumalis. Vancouverensis and origenes, though represented by f a i r l y adequate, samples cannot be separated from caurina. Boria i s l i k e l y best considered a variant of a c t u o s a . T h e samples of abieticola sierrae and humboldtensis are too small for adequate analysis. No specimens of atrata or kensiensis were examined. As a f i n a l check to the above conclusions, a l l samples were pooled, and treated with an analysis of variance. Significance beyond the .01 level was obtained for a l l measurements except condylo-basal length i n females, and the ratio canine width divided by bulla 40 -length, which was not tested. Wright's work (1951) indicated that a l l North American marten are conspecific. As a result of his work and the conclusions arrived at during the present study, the following classification of marten has been prepared. This classification i s as natural (phylo-genetic) as present knowledge permits. The basis for the decisions invalued i n i t s organization are discussed in Part I I . The presence of a question name indicates doubt concerning the actuality of a sub-species. Subspecies group americana f • Martes americana americana (Kerr 1792) Martes americana abietinoides Gray 1865 Martes americana atrata ? (Bangs 1897) Martes americana abieticola ? (Preble 1902) Martes americana brumalis (Bangs 1898) Martes americana actuosa (Osgood 1900) Martes,^americana kenaiensis ? ( E l l i o t 1903) Subspecies group caurina Martes americana caurina (Merriam 1890) Martes americana origenes ? (Rhoads 1902) Martes americana humboldtensis ? Grinnell and Dixon 1926 Martes americana sierrae ? Grinnell and Storer 1916 Martes americana vancouverensis ? Grinnell and Dixon 1926 Martes americana nesophila (Osgood 1901) - 41 -Marten appear te follow certain of the rules of geographic distribution l i s t e d by Mayr (1942). The caurina group of the humid western coast is darker in colouration than i s the americana group of the drier east (Gloger's Rule). Northern subspecies of the americana group(actuosa and brumalis) are larger than those to the south (Bergmann's Rule). I am unable to t e l l to what degree these size relationships are phylo-genetic, and to what degree a result of convergent adaptation. The following discussion (Part II) assumes the relationship to be largely phylogenetic. This assumption i s made with the knowledge that i t i s -f u l l y untested, and open to criticism. Only future work w i l l t e l l how well founded i t , and the following discussion based on i t , i s . . An a r t i f i c i a l key to marten skulls i s presented in Appendix B. PART II PHYLOGENY, Discussion: The literature pertaining to the phylogeny of the mustelids and the genus Martes i s sparse, scattered and often lacking in general agreement. As an indication of this lack of interest i n the group, Yeager (1941) points out that of the fur animals he considered only badger, otter and raccoon have received less attention than the wolver-ine, fisher and marten together. Two approaches have been used i n this study i n an attempt to determine something of the raci a l history of marten. These are: (1) ah examination of current interpretation of classification and.of palaeontological data and (2) an interpretation of certain palaeocological - 42 -data. (1) Classificatory and palaeohtological data: Mustelids have in the past been divided into from two (Cope 1882) to fifteen (Pocook 1921) subfamilies, with from four to eight being most usual. The subfamily Mustelinae, along with Lutrinae and Melinae however, have i n almost a l l cases been considered an objective entity. In a similar fashion, the genus martes has been long recog-nized as representing a unit separable from most others. The origin of the genus Martes i s not well understood. The situation i s the usual one of having innumerable small branches given off hear the base of the major splitting of a mammalian group.... In the. Mustelidae the situation is further complicated ... by the fact that numerous known Miocene and Pliocene Mustelids are seldom real forerunners of the l i v i n g genera, but in almost every genus appear to be a l l that i s known of a distinct phylum.... (Simpson 1945) The same author,also remarks that a l l pre-Pleistocene records may be viewed.with.doubt since they almost always use the generic name in a broader sense .than among recent mammals and often prove to be erroneous, when, the species became better known. Another d i f f i c u l t y encountered i s that palaeontologists have made the Mustelidae, and especially the genus Martes, a repository for a l l doubtful arctoid carnivores. Schlosser (1923) suggested grouping a l l of these into the subfamily Stenoplesictinae, but i t has not been acted upon to any great extent. To make one problem more confusing s t i l l , not only i s the origin of the group obscure, but i t i s poorly represented in the f o s s i l record (Scott 1937). The Mustelinids (the oldest of the Mustelids) f i r s t appeared - 43 -in the lower Oligocene, while the genus Martes (in the conservative sense) i s f i r s t recorded from the lower Pliocene of the old world, and the upper Pliocene of the new (Simpson 1945). Simpson (1947) believed that migration across the Bering land bridge occurred with varying degrees of intensity through a l l the Ceno-zuic subsequent to the Oligocene, excepting for the Recent. He states the appearance of the Mustelidae, Mustelinae and a l l of the Musteline genera common to North America and Eurasia is essentially simultaneous on the two land masses.... I t i s possible but very uncertain, that most of. the migration was from the old world to the new. He added that Martes probably crossed f i r s t i n the upper Pliocene, but that the direction i s very much in doubt. Pleistocene migrations were very extensive, and interglacial in the main. (Fig. 4.) The following genera have been suggested by various authors to l i e in the direct phylogeny of the genus Martes; Palaeogale, Bun-aelurus (Schlosser 1877-90), S i n i c t i s , P l i o n i c t i 3 t piesiogulo, Oligobunis, Amphictis (Pilgrim 1932) Mustelavus (Scott 1937). Species of Martes suggested as lying close to modern forms include i Martes kinseyi (Gidley 1927), Martes gazini (Hall 1931), Martes .. t.' diluviana (Hall 1936). Pilgrim (1932) considered that Martes and Mustela have been distinct since about the Stampian (middle Oligocene). "Martes i s , on the whole a survival of a primitive l i n e a g e — i t being most similar of a l l Recent forms to the hypothetical ancestor of the Mustelids." L i t t l e i s known_ regarding Pleistocene Mustelids, either in Britain (Reynolds 1912), Eurasia (Dubois 1928) or America. Hall (1936) F i g . 4. (to follow page 43.) Fig* 4. Distribution in time of the subfamilies (upper)' and musteline genera (lower) of mustelidae* Cross bars represent probable time of inter-continental migration. Largely after Simpson (1945) and Simpson (1947). 4--M U S T E L I N A E M E L L I V O R I N A E M E L I N A E " I CZZ1 M E P H I T I N A E L E P T A R C I N A E |.'?V:'V^r>::^Z^'U^\ L U T R I N A E I ' B f ^ f ^ f ^ f — ^ M i te'^a M U S T E L A V U S ~] PLESICTIS I f t 3 v # l B U N A E L U R U S 7J P A L A E O G A L E ESfl. ' -SII O L I G O B U N I S P R O M A R T E S P A R O L I G O B U N I S M E G A L I C t l S A E L U R O C Y O N j I S C H Y R I C T I S I K V g - y j P L E S I O G U L O P R O P U T O R I U S I 1 L A P H Y C T I S lay)>V^iU^51 P L I O N I C T I S I I P S E U D I C T I S fctf^:?^:;?!:.^^ B R A C H Y P S A L I S M I O M U S T E L A J S I N I C T I S my&>\ C E R N I C T I S l-".-WV| C A N I M A R T E S T R I G O N I C T I S LEGEND M U S T E L A M A R T E S C O L D W O R L D N E W W O R L D B O T H O L D & N E W W O R L D S G U L O G R I S O N G R I S O N E L L A L Y N C - E T T ^ ' - y v ^ . r ^ - y O D O N V O R M E L A C H A R R O N I A Z O R I L L A P O E C I L I C - Q T I S P O E C I L O G A L E E O C E N E O L I G O C E N E M I O C E N E . P L I O C E N E P L E I S T O C E N E R E C E N T - 44 has presented the most recent survey of American Pleistocene Mustelids, and almost a l l that immediately follows i s from his publication. Nine genera of Mustelids (so called) are known from Pleisto-cene deposits, two are extinct and Enhydra and Conepatus are not recorded> while Spilogale, Mustela, Mephitis, Lutra.Taxidea, Gulb, and Martes are. The following then are the known Pleistocene recordings of marten, as li s t e d by Hall. Martes americana:The sole ascription to this subspecies appears to be Allen's (1876) reference to that found in Baird's collection from Pennsylvania. It may or may not be i n re a l i t y a Recent deposit. Martes americana sierrae: Hall in this (1936) paper placed his previously described (1926) M. caurina nobil-is, a Pleistocene recovery from California in the synonomy of sierrae. He s t i l l considered i t to be of Pleistocene origin. Martes pennanti:Described by Allen (1876) and Brown (1908) from Pennsylvania and Arkansas. The latte r reference i s the only one known to ae of a recovery of a s t i l l l i v i n g Pleistocene Martes lying outside it s Recent range., Martes diluviana: Known from four specimens. Hall ascribed Gidley and Gazin's (1933) description of M. parapennanti to this species. Hall concludes that Osborne's (1925) reference to a marten from Rancho l a Brea was an erroneous identification of a weasel (Mustela). I have nowhere found an opinion as to when the separation of the genus Martes into the subgenera Pekania (fishers) and Martes (martens) - 45 -occurred. Since the genus arose during the lower Pliocene, and both subgenera are known from the Pleistocene, i t follows that the separa-tion must have occurred some time during the Pliocene. In summary then i t appears that the origin of the genus Martes i s not at a l l well understood; that i t arose sometime about the lower Pliocene, possibly in Eurasia, having been distinct from Mustela from the Oligocene; that migration back and forth between the old and new world took place from the time of i t s origin to the P l e i s -tocene; that the separation into i t s two component subgenera occurred about the Pliocene and that by the Pleistocene both subspecies groups of marten (americana and caurina) were present i n North America. (2) Palaeoecological data: Zoologists have generally been slow to realize that there exist certain palaeoecological methods useful in the determination of biotic histories. It i s well recognized that animal distribution in the present i s largely governed by the responses of plants to climate (Shelford 1935, Clements and Shelford 1939). That the same has held in the past has not been so clearly understood. Nevertheless "the outline of the history of animals must coincide with that of the plants on which a l l animals ultimately subsist" (Hulten 1937). Cain's (1944) text is the most recent exposition of the sub-ject and the student desiring more detail should turn to that book. Since so l i t t l e i s known about the former distribution of the marten, and since so much more is known about the former distribution of floras, and i f marten have in the past been restricted to one of - 46 -those floras, i t follows that much of the history of the marten is to he found i n the history of that f l o r a . I t i s upon this hypothesis that the following discussion is based. It i s well recognized that marten, and indeed almost a i l mem-bers of the subfamily Mustelinae are found only within communities of plants dominated by boreal conifers. If one plots the distribution of the coniferous west coast floras, the boreal forest and the lake forest together with that of the marten in North American, the close relationship between the two is apparent. I t i s equally apparent when the same i s done for Asia, but somewhat more obscure in the case of Europe. It i s less easy to demonstrate that marten have, since their origin in the Pliocene, been possessed of the same relationship. I t i s known, however, that marten are not recognized to have shown specific changes since the Pleistocene, and a l l f o s s i l recoveries are known from within their present range. Further, Pliocene Charronia from India are not known to be different from the forms now liv i n g there, (Flower and Lyddeker 1891). In opposition to this, however, i s the fact that not a l l f o s s i l Martes are known to have come from regions contemporan-eously boreal in nature, although at the same time the reverse has not yet been demonstrated. It is probably quite safe to conclude that for the Pleistocene at least, and very l i k e l y for much or a l l of the Pliocene; the demands, in terms of environment, of the martens ancestral to those now l i v i n g , were not very different from what they are today. I t seems unlikely that the tolerances of the components of animal communities - 47 -would change more than their vegetational substrata, and these i t is known have changed very l i t t l e (Arnold 1949, Cain 1944). Hooper (1949) has,-for example, pointed out that a l l genera of North American rodents can be divided into three groups, each referable to the three major ter-tiary floras of the continent. The present vegetation of North America is derived chiefly from three major floras* the so-oalled Arcto-tertiary Flora, the Madro-tertiary or Sierra Madrean Flora and the Neotropical-tertiary or Carib-bean Flora. Marten today, and in a l l probability in the past, inhabited the whole of the Arcto-tertiary Flora. This f l o r a appeared sometime between Cretaceous and Eo.cene times at high latitudes, migrating south-, ward with climatic deterioration u n t i l during Miocene and Pliocene times i t reached i t s approximate present position, forming the west coast conif-erous forests the boreal forest, the Lake forest and the eastern hardwood forest (Axelrod 1937 and 1939, Berry 1923, Braun 1947 and 1950, Chaney 1936, 1940, 1947, Chaney, Condit and Axelrod 1944, Mason 1947, Simpson 1943, Scharff 1911). While the large proportion of Mustelinids and the ancestors of marten probably migrated south with the Arcto-tertiary Flora i n Morth America and Eurasia, the true martens (subgenus Martes) are believed to have evolved in Eurasia (Rhoads 1902, Davis 1939, Dalquest 1948) and to have arrived in North America near the beginning of the Pleistocene by crossing the Bering land bridge, as did most characteristic American Pleistocene animals (Colbert 1942, Romer 1945). Rhoads (1902) implied an earlier migration for the americana subgroup than for the caurina, - 48 -on the basis of the latter*s close cranial relationship to Eurasian species. Davis (1939) likewise hypothesized an earlier migration for the americana group. There i s some slight evidence besides Rhoads' comparative work that this sequence actually occurred. No true marten are known from North America prior to the Pleistocene. The caurina group i s known from the Pleistocene of the west, and the americana from that of the east; primitive forms are assumed to l i e farther from the point of origin than more recent forms (Matthew 1915, Taylor 1937). I t seems li k e l y then on the basis of the somewhat ten-uous evidence now at hand that Martes americana arrived in North America from Eurasia early in the Pleistocene, and that the americana group preceded the caurina. The glaciation of the Pleistocene, especially the last (Wisconsin) advance, has been the primary factor determining present day subspeciation and distribution (Deevey 1949). It appears from analysis of pollen deposits and from geographic data that during the maximum of the last (Wisconsin-Vashon) glaciation, the forests occupied by the marten found refuge in (1) nunataks (2) south of the ice sheet, east of the Driftless area (3) south of the ice sheet i n the Rocky Mountains (4) suuth of the ice sheet in the Coast, Cascade and Sierra Nevada Ranges and (5) i n the unglaciated regions of Alaska. With the retreat of the Wisconsin-Vashon sheet, l i t t l e or no migration occurred from nunataks; the boreal and lake forests moved north from the southeastern refuge, the«;former north to Labrador and west to-near the Mackenzie River region; the Roeky Mountain forests moved north to - 49 -about northern B r i t i s h Columbia; the Pacific Coast floras moved north into Alaska; and the Alaskan boreal forest moved east to beyond the Mackenzie- and south into B r i t i s h Columbia (Adams 1902 and 1905, A r t i s t 1939, Braun 1928, 1947 and 1950, Brown 1938, Buell 1946, Cain 1944, Clements and Chaney 1937, Cooper 1931 and 1942, Darlington 1943, Davis 1946, Deevey 1943, F l i n t 1947, Fuller 1935, Gleason 1923, Halliday and Brown i943, Hansen 1938, 1939, i940, 1941, 1943 and 1947, Hansen and Packard 1949, Harshberger 1911, Hay 1902 and 1923, Hulten 1937, Krauss and Kent 1944, Mason 1947, Nichols 1935, Potzger 1946, Potzger and Wil-son 1941, Raup 1941 and 1947, Rosendahl 1948, Sears -1935 a and b, 1938, 1942 and 1948, Transeau 1903 and 1905, Yoss 1934, 1937 and 1939, Wilson 1938, WvflHe-Edwards 1937). It i s undoubtedly true that marten likewise spent the Wisconsin-Vashon glaciation in a l l of the above l i s t e d refugia, and followed the migrations of their forest environment to their present position. E v i -dence for this conclusion is as follows; (a) In both the old and the. new world marten have . . adapted themselves to the whole of the range of. coniferous l i f e form. This indicates that the gl a c i a l refugia of marten must have been many and large (since l i t t l e depauperization of "biotypes" has occurred) and i n fact must have been precisely those of the forest i t now lives i n . (b) Marten today s t i l l have parts of their range . . extending into a l l of the several gl a c i a l refugia supposedly occupied by the forest they now live i n . (Fig. 5). (c) There i s some slight indication that greatest . . densities of marten are found near the regions of former glacial refugua. One would expect this to be the case i f these refugia had in ' F i g . 5. 1 (to follow page 49.) Fig* 5. Areas of the present day distribution of marten lying outside the range of Pleistocene glaciation. Dotted areas mark marten distribution; solid line represents the boundary of maximum Wisconsin glaciation; dotted lines indicate older glaciations; lines A to E repres-ent trend of post-Wisconsin gla c i a l retreat. I t i s now believed that Newfoundland underwent complete glac-iation during the Wisconsin. After Antevs (1928, 1929, and 1931), F l i n t (1945 and 1947), and Thwaites (1934). - 50 -the past been centers of postglacial dispersal, and i f the subsequently acquired areas were s t i l l i n the process of being populated. This same relationship has been found to occur i n the density of forest trees in Canada (Halliday and Brown 1937). See F i g . 5. We are now in a position to speculate upon the postglacial history of the marten. Since Martes americana americana s t i l l ranges south into south eastern refugia, i t doubtless spent the Wisconsin Ice Age there as did brumalis, and subsequently extended i t s range north to i t s present range l i m i t . Abietinoides and origenes s t i l l exist in part in the Rocky Mountain refugium, and have since moved north to form their present distribution. Origenes has not arrived north of the g l a c i a l boundary, and has not yet made contact with caurina from which i t was previously isolated. Caurina. humboldtensis and sierrae s t i l l range into the coastal refugium. Of these only caurina moved north with the retreat of the Vashon sheet. Actuosa and kenaiensis s t i l l occur in the Yukon refugiumjonly actuosa has migrated eastward, extending i t s range beyond the Mackenzie and into -B r i t i s h Columbia. Abieticola, whose distribution l i e s in the center of the boreal forest, does not range intovany of the g l a c i a l refugia, and i t s relationship to these i s more d i f f i c u l t to determine. I t seems l i k e l y , however, that abieticola represents an intergradation between actuosa and americana which has resulted from mixing of two strains on meeting after postglacial migration. The migrations des-cribed above are mapped on F i g . 7. It i s important here to take careful notice of the fact that a l l of the marten migrations described above correspond very closely F i g . 6. (To follow page 50.) F i g . 6. Average number of marten trapped per 1000 square miles of forested land per state or province per year. Figures derived from Anon. (1927-1950). Anon (1950), Minville (1946), Ullman (1949) and Yeager (1950) . The number of years considered varies from two (Maine) to thirty-two (Quebec). - 51 -in terms of direction and distance travelled, with the migrations of coniferous forest species moving out of the same refugia, as determined hy areographic and pollen analytical methods described by the students li s t e d above. Insofar as these studies are accurate, the above doubt-less represents a f a i r l y precise picture of the postglacial migration of the marten in north America, and probably of eertain other boreal animals as well. Assuming the above hypotheses to be reasonably sound, there are certain other phylogenetic relationships that can be considered. It w i l l be noticed that a l l marten of the south eastern and Yukon ref-ugia are of the americana group, while those of the Pacific Coast refugium are caurina. In the Rocky Mountain refuge however, both groups are represented (abietinoides and origenes). It follows that origenes and caurina were separated prior to or by the Vashon (Wisconsin) sheet and that the separation into americana and caurina subspecies groups must have occurred long before Vashon time, probably, as has been suggested, very early i n the Pleistocene and in Eurasia. It w i l l have likewise been observed from the data presented in Part I of this study that M. a. actuosa and M. a brumalis were more similar to each other than they were to any others, and that abietinoides and americana were also similar. If these similarities can be consid-ered to possess a phylogenetic relationship, i t implies thats (1) the separation of brumalis from actuosa occurred during Wisconsin time at the.latest, while they were isolated-from each other; that (2) the same is true for americana and abietinoides; that (3) abieticola has arisen F i g . 7. (To follow page 51.) F i g . 7. Assumed post-Wisconsin marten migrations. Those subspecies not marked by an arrow are believed to have moved l i t t l e or not at a l l from their location during the Wisconsin maximum. - 52 -since Wisconsin time through the interbreeding of actuosa and americana; and that (-4) actuosa-brumalis must have been separated from abietinoides-americana sometime between the early Pleistocene at the earliest arid the Illinoian glacial advance at the latest. The insular subspecies of marten have yet to be considered. Atrata has not been examined but i s said by Rhoads (1902) to be closer to americana than to brumalis; atrata probably arose from americana either during or subsequent to the Wisconsin. Vancouverensis and nesophila have doubtless sprung from caurina. Vancouver Island is con-sidered to have been glaciated (Flint 1947) and doubtless migration to the island and subspeciation has occurred since Vashon-Wlsconsin time. Nesophila shows far greater differences from caurina than does vancou-verensis and this may be due to any or a l l of several facts. It has been suggested that the Queen Charlotte Islands have escaped glaciation (McCabe and Cowan 1945). This would mean that nesophila has been iso-lated for a much longer time than has vancouverensis. Small insular populations undergo a change in the frequency of their gene- distribution such as to result in a homogenous population with phenotypes nonadapt-ively different from those of the parent stock (Huxley 1940). Whatever the factor at work i s , i t i s not possible to determine more of the origin of nesophila than to al l y i t with caurina. The foregoing may be summarized as follows: there i s a certain amount of evidence to show that pre-americana crossed to the new world sometime between the late Pliocene and the Kansan, while pre-caurina probably followed at a later date, but also not later than Kansan. - 53 Between the Kansan and the i l l i n o i a n americana had divided into abietinoides-americana and brumalis actuosa* Prior to the Wisconsin, origenes and caurina were separated. The advance of the Wisconsin sheet saw a segregation of marten in four refugia, each isolated from the other, during which time or subsequently subspeciation occurred. Of these refugia the south eastern repopulated most of Canada east of the Rockies (brumalis, americana, probably atrata and abieticola in part); the Pacific and Rocky mountain refugia Canada west of the Rockies (caurina, origenes, vancouverensis and abietinoides together with humboldtensis and sierrae); and the Alaska refuge the Alaskan and Canadian northwest (actuosa, kenaiensis and abieticola i n part). M. nesophila i s of a caurina origin, but i t s history i s poorly under-stood. Conclusions; On the basis of the comparison of skulls described in Part I, and of the discussion of probable history of the marten in Part II, a provisional phylogeny of the subgenus has been prepared, and is presen-ted in F i g . 8. This phylogeny assumes that a l l subspecies of marten discussed i n this paper are valid ones, which i s of course, not necessar-i l y the case. SUMMARY 1. A s t a t i s t i c a l analysis of 418 skulls of North American marten of the subspecies brumalis, americana, abieticola, actuosa, Fig., 8. (To follow page 53.) F i g . 8. A hypothetical phylogeny of the martens of North America. V A N C O U V E R E N S I S A T R A T A A B I E T I N O I D E S A M E R I C A N A C A U R I N A S I E R R A E N E S O P H I L A O R I G E N E S H U M B O L D T E N S I S CAURINA GROUP \ A B I E T I C O L A ? N \ \ A C T U O S A B R U M A L I S K E N A I E N S I S ? AMERICANA GROUP - 54 -abietinoides. caurina. oro^enes, sierrae, humboldtensis, vancouverensis, and-nesophila indicates that these, plus probably those not studied (atrata and kenaiensis) f a l l into two natural groups, the americana group (formerly the species americana) and the caurina group (formerly the species caurina). The unpublished work of Wright indicates that these two groups are conspecific, and that only one species, comprised of these two groups, exists on the continent. 2. What is probably adequate evidence for the subspecific status of brumalis, americana, actuosa, caurina and nesophila is presented. Abietinoides i s , on the basis of the characters studied, probably separable from americana, while vaneouverensis, caurina and origenes are inseparable. A l l three must for the time being be con-sidered distinct however, u n t i l further examination of them is made on the basis of their geographic distribution. Abieticola i s not well represented in the samples studied. I t possibly represents a population of intergrades of actuosa and americana.Sierrae and humboldtensis are represented by samples too small for analysis. 3. An a r t i f i c i a l key to the skulls of marten of North America is; presented. 4. The history of martens i s not well known. The genus Martes probably arose in the lower Pliocene, possibly in Eurasia, having been distinct from Mustela since the Oligocene. Migration between the old and new world occurred from, the time of i t s origin to the Pleistocene. Separation into the subgenera of fishers and martens probably occurred during the Pliocene. - 55 -5. Pre-americana and pre-caurina both appear to have had their origin i n the old world, the former preceding the latter i n i t s migration to the new, both arriving sometime between the late Pliocene and the Kansan. Between the Kansan and the Il l i n o i a n pre-americana had segregated into- abi-etinoides-americana and brumalis-actuosa. Prior to the Wisconsin origenes and caurina were separated. The advance of the Wisconsin ice sheet resulted in a gathering of marten into four refugia, during which time or subsequently, f i n a l subspeciation occurred. Repopulation of North America following the f i n a l ice retreat originated by migration from these refugiaj atrata, brumalis, americana and abieti-cola in part from a south eastern refuge, origenes -and abietinoides from one in the southern Rocky Mountains, humboldtensis, sierrae and caurina from one i n the southern coast range, and actuosa and kenaiensis. from an Alaskan refuge. Nesophila and "vangouverensis are related to caurina, the former much less closely than the la t t e r . 6. A hypothetical -phylogeny of North America is diagrammed. LITERATURE CITED Adams, C. C., 1902, Postglacial origin and migrations of the l i f e of the north-eastern United States. Journ. G-eogr. Is 303-310, 352-357; 1905 The post-glacial dispersal of the North American biota. B i o l . B u l l . 9 : 53-71. Allen, G. M., 1904 Check l i s t of the mammals of New England. Occ. papers Boston. Soc. Nat. Hist. 7; 1-35. 1942 Extinct and vanishing mammals of the western hemisphere, with the marine species of a l l the oceans. Lancaster, Amer. Comm. Internat. Wildl. Protection, Spec. Publ. 11, pp. xv 620 i l l u s . Allen, J. A., 1869 Catalogue of the mammals of Massachussetts; with a critical.,revision, of, the, species. B u l l . Mus. Comp.. Zool. Is. 161-167. 1876 Geographic variation among North American mammals, especially in. respect to size. B u l l . U. S. Geol. and Geogr. Sur,v. territories. 2s 309-344. 1895 On the names of mammals given by Kerrmin his "Animal Kingdom", published i n 1792. B u l l . Amer. Mus,. Nat. Hist. 7s 179-192. Alston, E.R., 1879 On the specific identity of the B r i t i s h martens. Proc. Zool. Soc. London, 1879: 468-474. Ames, A. E., 1874 Mammalia of Minnesota. B u l l . Minn. Acad. Nat. Sci. 1874s 68-71. Anderson, R. M. 1934a Mammals of the Eastern Arctic and Hudson Bay. In: Canada's Eastern Arctic, Cndn. Dept. Interior Publ., pp. 67rl08. - 57 LITERATURE CITED continued Anderson, R. M., 1934b 1937, 1946, The distribution, abundance and economic importance of the game and fur-bearing mammals of western North America. Proc. 5th Pac. S c i . Congr. 1933, 5; 4055-4073. Mammals and birds._ In: Canada's Western Northland, Cndn. Dept. Mines Resources, pp. 97-122. Catalogue of Canadian Recent mammals. B u l l . Nat. -'Mus. Can. 102, pp. v i , plus 238. Anonymous, 1927-1950 Fur production of Canada, Annual reports, Cndn. Dept Trade Comm. 1950, Antevs, E., 1928, 1931, The Canada Year Book, 1950; the o f f i c i a l s t a t i s t i c a l annual of the resources, history, institutions, and social and economic conditions of Canada. Cndn. Dept. Trade Comm., pp. XLIV plus 1238 i l l u s . The ls,st glaciation with special reference to the ice retreat i n north-eastern North America, Amer. Geogr. Soc. Res. Ser. 17, pp. X plus 292 i l l u s . Late-glacial correlations and ice recession in Manitoba. Mem. Geol. Surv. Can. 168, pp. YT plus 76 i l l u s . Anthony, H. E., 1928, Fieldbook of North American mammals; descriptions of every mammal known north of the Rio Grande, together with brief accounts of habits, geographical ranges, etc., New York, Putnams, pp. XXVI plus 625 plus 48 plates, i l l u s . Arnold, C. A., 1947. An introduction to paleobotany, New York, McGraw-Hill, pp. XI plus 433 i l l u s . - 58 -LITERATURE CITED continued • i A r t i s t , R. C , 1939, Pollen spectrum studies on the Anoka sand plain in Minnesota. Ecol. Monogr. 9t 493-536. Audubon, J. J., and J. Bachman, .1856, The. quadrupeds of North America, New York, V. G. Audubon, pp. VI plus 348 i l l u s . Axel rod, D. I., .1937, A Pliocene f l o r a from the Mount Eden beds, Southern California, Carnegie Inst. Wash. Publ. 476, pp. 125-183. 1939, A Miocene f l o r a from the western border of the Mohave Desert. Carnegie Inst. Wash. Publ. 516, pp. 1. - 129 i l l u s . Bailey, V., 1936, The mammals and l i f e zones of Oregon, N. Amer. Fauna 55 : 1-416 i l l u s . Baird, S. F., 1857. General report upon the mammals of the several Pacific railroad routes. U.S.Pac. R.R. Expl. Surv. 8. 1., pp. XXXIV plus 764 plus 60'plates 1859, Mammals of North America; the descriptions of species based chiefly on the collection in the museum of the Smithsonian Institution, Phila., Lippincott, pp. XXXIV plus 764 i l l u s . Bangs, 0., 1897, Preliminary description of the Newfoundland marten. Amer. Nat. 31 : 161-162. 1898, A l i s t of the mammals of Labrador. Amer. .Nat. 32 : 489-507. - 59 -LITERATURE CITED continued Berry, E. W., 1923, Tree ancestors; a glimpse into the past. Baltimore, Williams and Wilkins, pp. VI plus 270 i l l u s . Bowman, T., Forest physiography, New York, Wiley, pp. XXII plus 1911, 730 i l l u s . , and maps. Brandt, J . F•, . 1855, Beitrage zur nahern Kenntniss der Saugethiere Russlands. Mem. Math., Pliys., et Naturelles de l'Acad. Sci;- St. Petersburg, 7 : 2-24. Braun, E. L«, 1928, Glacial and post-glacial plant migrations indicated by r e l i c colonies of southern Ohio. Ecol.--9: 284-302. 1947, Development of, the deciduous forests of eastern North America. Ecol. Monogr. 17: 211-219. 1950, Deciduous forests of eastern North America. Phila., Blakiston, pp. XIV plus 596 i l l u s . Brongersma, L.D., 1941, On the remains of carnivora from cave deposits in Java and Sumatra with notes on Recent species. Zool. Med.Keri.den, 23 : 114-147. Brown, B., 1908, The Gonard Fissure, A Pleistocene bone deposit in northern Arkansas: with descriptions of two new genera and twenty new species of mammals. Mem.Amer Mus. Nat. Hist. 9: 155-208. Brown, C. A., 1938, The f l o r a of the Pleistocene deposits in the Western Florida Parishes; West Feliciana Parish and East Baton Rough Parish, Lousiana. La. Dept. Cons, B u l l . 12; 59-94. - 60 LITERATURE CITED continued Buell, M. F., 1946, The age of Jerome Bog, "A Carolina Bay". S c i . 103: 14-15. Buffon, G. L., and L. J. M. D'Aubenton, 1758 . Histoire naturelle, generale et particuliere, avec and 1765, l a description du Cabinet du Roi. Vol. 7, 1758; vol. 13, 1765, Paris, L'Imprimerie Royale. Cain, S. A., 1944, Foundations of plant geography. New York, Harper, pp. XVI plus 556 i l l u s . Chaney, R. W., 1936, Plant distribution as a guide to age determination. Journ. Wash. Acad. S c i . 26: 313-324. 1940, Tertiary forests and continental history. B u l l . Geol. Soc. Amer. 51. 469-488. 1947, Tertiary centers and migration routes. Ecol. Monogr, 17: 139-148. Chaney, R. W.; C. Condit, and D. I. Axelrod, 1944, Pliocene floras of California. Carnegie Inst. Wash. Publ. 553, pp. V plus 407 i l l u s . Clements, F• E., and R. W. Chaney, 1937, Environment and l i f e in the Great Plains. "Carnegie Inst. Wash. Suppl. Publ. 24 (Rev. Edit.), pp. 1-54. Clements, F.E., and V. E. Shelford, 1939,, Bio-ecology. New York, Wiley, pp. VIII plus 425 i l l u s . Colbert, 1942, The geologic succession of the Proboscidia. In: The Proboscidia, by H. F. Osborn, New York, Amer. Mus. Nat. Hist., pp. 1421-1521. - 61 LITERATURE CITED continued Cooper, W. S., 1931, Third expedition to Glacier Bay, Alaska. Ecol. 12s 61-95. . 1942, Vegetation of the Prince William Sound region, Alaska; with a brief excursion into post-Pleisto-cene climatic history. Ecol. Monogr. 12: 1-22. Cope, E. D., 1882, On the systematic relations of the Carnivera F i s s i p e d i a . P r o c . Amer. Philos. Soc. 20: 471-475; also Palaeont. B u l l . 35: 471-475. Cory, C. B., 1912, The mammals of I l l i n o i s and Wisconsin. Publ. F i e l d Mus. Nat..Hist. 153: 1-505. Coues, E., 1877, Fur-bearing mammals; a monograph of North American mustelidae. U. S. D. I. Geol. Surv. Misc. Publ. 8, pp. XIV plus 348 i l l u s . . Cuvier, Q. L. C. F. D., 1823, Martes des Hurons, Mustela-huro. In: Dictionnaire des sciences naturelles, v o l . 29, p. 256, Strasbourg-Paris, Levrault. Dalquest, W. W., 1948, Mammals of Washington. Publ. Univ. Kansas 2; 1-444. Darlington, H. C , 1943, Vegetation and substrate of Cranberry Glades, West Virginia. Bot. Gaz. 104: 371-393. Davis, J. H., 1946, The peat deposites of Florida. B u l l . F l a . Geol. Surv. 30: 1-247. Davis, W. B., 1939, The Recent mammals of Idaho, Caldwell, Caxton, pp. 400 i l l u s . - 62 -LITERATURE CITED continued Deevey, E. S. Jr., 1943, Additional pollen analyses from southern New England. Amer. Journ. Sci. 241: 717-752. 1949, Biogeography of the Pleistocene. B u l l . Geol. Soc. Amer. 60: 1315-1416. De Kay, J. E., . 1842, Zoology of New-York, or the New-York fauna.... Part 1, Mammalia. Albany, Thurlow and Weed, pp. XV plus 146 i l l u s . de Voss, A., 1951, Recent findings in fisher and marten ©fioiogy and management, trans. 16th. N. Amer. Wildl. Conf., pp. 498-507. Downing, S. C , 1948, A provisional check-list of the mammals of Ontario. R.O.M.Z. (Toronto) Misc. Publ. 2: 1-11. Dubois, G, ' ~ 1928, Marte de l a Tourbe de Marquion.Ann. soc. geol. L i l l e , 52: 280-285. Edwards, R. Y., 1950, Variations in the fur productivity of northern B r i t i s h Columbia in relation to some environmental factors. Unpubl. M. A.. Thesis, Univ. B. C , pp. 126. E l l i o t , D. G., 1901, & synopsis of the mammals of North America and the adjacent seas. Publ. F i e l d Columbian Mus. 45: 1-471. 1903, Description of an apparently new sub-species of marten from the Kenai Peninsula, Alaska. Publ. F i e l d Columbian Mus. 72, (9): 151-152. - 63 continued LITERATURE CITED E l l i o t , D. . 1905a, 1905b, A check l i s t of mammals of the North American continent, the West Indies, and the neighboring seas. Publ. F i e l d Columbian Mus. Zool Ser. 6, pp. IV plus 761 i l l u s . Descriptions of apparently new mammals of the genera Qvibos, Cyhomys and Mustela. Proc. B i o l . Soc. Wash. 18: 135-140 Emmons, E., 1840, Report on the quadrupeds of Massachusetts. Cambridge Univ. Press, pp. 86. F l i n t , R. F., (ieditor), 1945, Glacial map of North America. Geol. Soc. Amer. Spec. Paper 60> pp. VIII plus 37 plus enclosure. 1947, Glacial geology and the Pleistocene epoch. New York,. Wiley,., pp.. XVIII plus 589 plus maps, i l l u s . Flower, W. H., and R. Lydekker, 1891, An_ introduction to the study of mammals livin g and extinct. London, Black, pp. XVT plus 763 i l l u s . Forster, J. R., .1772, Account of several quadrupeds from Hudson's Bay, Philos. trans. Royal Soc. London 62: 370-381. Frisch, J. L., . 1775, Das Natur-system der vierfuszigen Thiere, in Tabellen, darinnen alle Ordnungen, Geschlechte und Arten, nicht nur rait bestiramenden Benenn-ungen, sondern beygeseszten unter-scheidenden Kennzeichen angezeigt werden, zum Nutzen der erwachsenen Schuljugend. Glogan, Gunther, pp. 35 i l l u s . Fuller, G. D., 1935, Postglacial vegetation of the Lake Michigan region. Ecol. 16: 473-487. - 64 -LITERATURE CITED continued Gidley, J. W., 1927, A true marten from the Madison Valley (Miocene) of Montana. Journ. Mammal. 8: 239-242. Gidley, J. W., and C. L. Gazin, 1933, New Mammalia in the Pleistocene fauna of the Cumberland Cave. Journ. Mammal. 14: 343-357. Gilpin, J. B., . 1860, On the mammals of Nova Scotia, no. III. Trans. Nova Scotia In3t. Nat. S c i . 3: 11-14,. Gist e l , J., . 1848, Handbuch Naturges. Stuttgart, pp. ? Gleason, H. A., 1923, Vegetational history of the middle west. Ann. Assoc. Amer. Geogr. 12: 39-85. Gray, J. E., 1843, List of the specimens of Mammalia in the collection of the B r i t i s h Museum. B r i t . Mus. Nat. Hist., pp. XXVIII plus 216. 1865, Revision of the genera and species of Mustelidae contained in the B r i t i s h Museum. Proc. Zool. Soc. London, 1865: 100-154. G r i f f i t h , E., C. H. Smith, and E. Pidgeon. 1827, Cuvier's Animal Kingdom, vol. 2, The Class Mammalia, London, Whittaker, pp. I plus 513 i l l u s . Grinnell, J., and J. Dixon, 1926, . Two new races of the pine marten" from the Pacific - coast of North America. Univ. C a l i f . Publ. Zool. 21: 411-417. - 6 5 -LITERATURE CIxED continued Grinnell, J., J. S. Dixon, and J. IS. Linsdale, 1937, Fur-bearing mammals of California, their natural history, systematic status, and relations to mani Univ. C a l i f . Pres3, 2 vols. Grinnell, J., and T. I. Storer. 1916, Diagnoses of seven new mammals from east-central California. Univ. C a l i f . Publ. Zool. 17: 1-8 Hall, S. R., 1926, A new marten from the Pleistocene cave deposits of California. Journ. Mammal. 7: 127-130. 1931, Description of a new mustelid from .the later Tertiary of Oregon, with assignment ;• of Par ictus primaevus .to the Canidae. Journ. Mammal. 12: 156-158. 1936, Mustelid mammals from the Pleistocene of North America, with systematic notes on some Recent members of the genera Mustela, Taxidea and Mephitis. Carnegie Inst. Wash. Publ 473,pp. 41-119. 1946, Mammals of Nevada, Univ. C a l i f . Press, pp. VI plus 710 i l l u s . Halliday, W. E. D., 1937 A forest classification for Canada. Cndn. Dept. Mines Resources, Forest Serv. B u l l . 89, pp..50 plus map. Halliday, W. E. D., and A. W. A. Brown, 1943, The distribution of some important forest trees in Canada. Ecol. 24: 353-373. Hamilton, W. J., Jr., ^ 1943. The mammals of eastern United States; an account of Recent land mammals occurring eaet of the Mi3 sissippi. Ithaca, Comstock, pp. 432 i l l u s . - 66 -LITERATURE CITED continued Hansen, H. P., . . 1938, 1939, 1940, 1941, 1943, 1947, Postglacial forest succession and climate i n the Puget Sound region. Ecol. 19: 528-542. Postglacial vegetation of the Driftless Area of Wisconsin. Amer. Midi. Nat. 21: 752-762. Paleoecology^of two peat bpgs in southwestern B r i t i s h Columbia. Amer. Journ. Botany 27: 144-149. Further study of post-Pleistocene bogs in the Puget lowland of Washington. B u l l . Torrey-Bot. Club 68: 133-148. A pollen study of a subalpine bog in the Blue Mountains of northeastern Oregon. E.coli, 24: 70-78. Postglacial forest succession, climate, and chronology in the Pacific northwest. Trans. American Philos. Soc. n. s. 37: 1-130. Hansen, H. P., amd E. L. Packard, 1949, Pollen analysis and the age,of proboscidian bones near Silverton, Oregon. Ecol. 30: 461-468. Hantzch, B., 1909. Contributions to the knowledge of extreme north-eastern Labrador, transl. by M. B. A. Anderson, Cndn. F i d . Nat. 46: 34-36 (1932). Harshberger, J. W., 1911, Phytogeographic survey of North America. New York, Stechert, pp. Hay, 0. P., 1902 Bibliography and catalogue of the f o s s i l Vertebrata of North America. B u l l . .U.S. Geol. Surv. 179, pp. I l l plus 868. 1923, The Pleistocene of North America and i t s veriebrated animals from the states east of the Mississippi River and from the Canadian provinces east of longitude 95 • Carnegie Inst. Wash. Publ. 322, pp. VIII plus 499 i l l u s . - 67 -LITERATURE CITED continued Hershkovitz, P., 1948, Names of mammals dated from Frisch, 1775, and Zimmerman, 1777. Journ. Mammal. 29s 272-277. Hooper, E. T., 1949, Faunal relationships of Recent North American rodents. Misc. Publ. Mus. Zool. Univ. Mich. 72, pp. 1-28. Hulten E., 1937, Outline of the history of Arctic and Boreal biota during the Quaternary period. Their, evolution during and after the gl a c i a l period as indicated by the equiformal progressive areas of present plant species. Stockholm, Bokforlags Thule, pp. 168 i l l u s . Huxley, J., (editor), . 1940, The new systematics, Oxford Univ. Press, pp. VIII plus 583 i l l u s . Jurgenson, P.B., 1939, Types of habitat and the forest marten; a contribution to the ecology of the marten. Problems Ecol. and Biocenology Leningrad 4s 142-173. Kaup, J. J., 1829, Skizzirte Entwickelungs-Geschichte und naturliches System der Europ-iaischen Thierwelt, vol. 1., Darmstadt, Weber, pp. Kerr, R., 1792, The animal kingdom, or zoological system of the cele-brated S i r Charles Linnaeus. Class I, Mammalia ... being a translation, of that part of the Systema Naturae as lately published, with great improvements.... 1 London, Murray and Faulder, pp. XII plus 400 i l l u s . Krauss,:R. w., and 0. N. Kent., 1944, Analyses and correlations of four New Hampshire bogs. Ohio Journ. S c i . 44: 11-17. - 68 -LITERATURE CITED continued i Kuhl, H. von, 1820, Beitrage zur Zoologie und vergleichenden Anatomie, (Zweite Abtheilung). Frankfurt, Hanan, pp. ?. Linnaeus, C , . 1758, Systema naturae per regna t r i a naturae, secundum classes, ordines., genera, -species cum characteribus, .differentiis, synonymis, l e c i s j editio decima, reformata. Stockholm, Laurentii S a l v i i , vol. 1, pp II x 824. Marshall, W. H., 1951, An age determination method for the pine marten, Journ. Wildl. Man. 15s 276-283. Mason, H. L., 1947, Evolution of certain f l o r i s t i c associations i n western North America. E.c.ol. Monogr. 17$ 201-210. Matthew, W. D., 1915, Climate and evolution. Ann. N. Y. Acad. S c i . 24: 171-318. Mayr, E., 1942, Systematics and the origin of species from the view-point of a zoologist. New York, Columbia Univ. Press, pp. XIV plus 334 i l l u s . McCabe, T. T., and l.McT. Cowan, : 1945, Peromyscus maniculatus macrorhinus and the problem of insularity. Trans. Royal Cndn. Inst. 1945: 117-215. Merriam, C. H., 1890, Descriptions of twenty-six new species of North American Mammals. N. Amer. Fauna 4, pp. V plus 54 i l l u s . M i ller, G. S., 1900, Key to the" land mammals of northeastern North America. B u l l . N. Y. State Mus. 8: 59-160. - 69 -continued LITERATURE CITED Mill e r , G. 5. 1912a 1912b, 1924, Catalogue of the mammals of western Europe (Europe exclusive of Russia) in the collection of the Bri t i s h Museum. B r i t . Mus. Nat. Hist., pp. XV plus 1019 i l l u s . L i s t of North American land mammals in the United States National Museum 1911. B u l l . U. S. Nat. Mus. 79, pp. XIV plus 455. L i s t of North American Recent, mammals 1923. B u l l . U. S. Nat. Mus. 128, pp. XVI plus 673. " Miller, 0. S., and J . A. G. Rehn, 1901, Systematic results of the study of N. A. land mammals to the year 1900. 'Proc. Boston Soc. Nat. Hist. 30: 1-352. Minville, E., (editor), 1946, Peche et chasse. Montreal, Ecole des Etudes com-mercials, pp. 580. Morris, R. F., 1948, The land mammals of New Brunswick. Journ. Mammal. 29: 165-176. Munro, J. A., and I. MoT. Cowan, 1947, A review of the bird fauna of Br i t i s h Columbia. B. C. Prov. Mus. Spec. Publ. 2, pp. 285 i l l u s . Newsom, W. M., 1937, Mammals of Anticosti Island, Journ. Mammal. 18: 435-442. Nichols, G. E., 1935, The hemlock - white pine - northern hardwood region of eastern North America. Ecol. 16: 4034422. - 70 — LITERATURE CITED continued Nilsson, S., 1820, Skandinavisk Fauna. En Handbok for Jagare och Zoologer. Forsta Delen. Lund, Tryckt u t i Berlingska Boktry-ckereit, vol. 1. Ognev, S. I., < 1928, New data on the systematic and the geographical distribution of some species of the family Mustel-idae. Mem. Zool. Soc. Amis. S c i . nat. Moscow, 2: 1-30. 1928- The mammals of eastern Europe and northern Asia, 1950 Vol. 2 (1931) and vol. 3 (1935), Carnivora, Moscow and Leningrad, Publ. Acad.. S c i . U. S. S. R. Okland, F., . 1937, Die geographischen Rassen der extra-marinen Wirbel-tiere Europas. Zoo-geographica 3s 389-484. Osborn, H. F., 1925, Mammals and birds of the California tar pools. Nat. Hist. 25s 527-543. Osgood, ¥. H., 1900, 1901, Biological reconnaissance of the Yukon River region ... annotated l i s t of the mammals. N. Amer. Fauna 19, pp. 1-43. Natural history of the Queen Charlotte Islands, B r i t i s h Columbia. Natural history of the Cook Inlet region, Alaska. N. Amer. Fauna 21, pp. 1-87. Pallas, P. S., 1776-1780 (exact date unknown), Spicilegia zoologica, quibis novae imprimis et obscurae animalium species iconibus descriptionibus atque commentariis illustrantur, v o l . 2 Berlin. Palmer, T. S., . 1904, Index Generum Mammaliums a l i s t of the genera and families of mammals. N. Amer. Fauna 23, pp. 1-984. - 71 -LITERATURE CITED continued Pilgrim, G. E., 1932, Gn the genera Trochictis, Enhydrictis and Trocharion with remarks on the taxonomy of the mustelidae. Proc. Zool. Soc. London 1932t 845-867. Pinel, M., 1792, Recherches sur une nouvelle methode de classification des quadrupeds, fondee sur l a structure mechanique des parties osseuses qui servent a 1'articulation de l a machoire inferieure. Actes Societe d'histoire natu r e l l e 3 de Paris Is 50-72. Pocock, R. I., 1921, On the external characters and classification of the Mustelidae. Proc. Zool. Soc. London 1921: 803-837. Ponomarev, A. L., 1946, Kidas. Bull . Soc. Nat. Moscow, Sect. B i o l . , 51: 79-83. Poole, A. J., and V. S. Schantz, 1942, Catalogue of the type specimens of mammals in the , United States National Museum, including the Biologi-cal Survey's collection. B u l l . U. S. Nat. Mus. 178, pp. XIII plus 705. Potzger, J. E., .1946, Phytosociology of the primeval forest in central-northern Wisconsin and upper Michigan, and a brief post-glacial history of the Lake Forest formation. Ecol. Monogr. 16: 211-250. Potzger, J. E., and J. T. Wilson, .1941, Post-Pleistocene forest migrations as indicated by sediments from three deep inland lakes. Amer. Midi. Nat. 25: 270-289. Preble, E. A., 1902, A biological investigation of the Hudson Bay region. N. Amer. Fauna 22, pp. 1-140. - 72 -LITERATURE CITED continued Preble, E. A., 1908, A biological investigation of the Athabaska - Mackenzie region. N. Amer. Fauna 27, pp. 1-574. Rafinesque, C. S., 1819, Description of a new species of North American marten (Mustela vulpina). Amer. Journ. S c i . Is 82-84. Raup, H. M., . 1941, Botanical problems in boreal America. Bot. Rev. 7: 147-248. 1947, Some natural f l o r i s t i o areas i n boreal America. Ecol. Monogr. 17s 221-234. Reynolds, S. H., 1912,. A monograph of the Br i t i s h Pleistocene Mammalis, vol. 2,.; pt. 4, Mustelidae. Palaeont. Soc. London Publ., pp. 256 i l l u s . Rhoads, S. N., 1902, 1903, Synopsis of the American martens. Proc. Acad. Nat. S c i . Phila. 54s 443-460. The mammals of Pennsylvania and New Jersey. Phila. privately printed, pp. 266 i l l u s . Richardson, J., 1829, Fauna-boreali americana; or the zoology of the northern parts of B r i t i s h America, vol. 1, quadrupeds> London, Murray, pp. XLVI plus 300- i l l u s . Romer, A. S., 1945, Vertebrate paleontology, 2nd. edit., Univ. Chicago Press, pp. 687 i l l u s . Rosendahl, C. 0., 1948, A contribution to the knowledge of the Pleistocene f l o r a of Minnesota. Ecol. 29: 284-315. - 73 -LITERATURE CITED continued Ru3t, H. J., .1946, Mammals of Northern Idaho. 308-327. Journ. Mammal. 27: Sanborn, C. C , 1947, Catalogue of type specimens of mammals in Chicago Natural History Museum. Chicago Nat. Hist. Mus. Sci. Ser. 32: 205-293. Scharff, R. Fv 1911, Distribution and origin of l i f e in America, London, Constable, pp. XVI plus 497 i l l u s . plus maps.. Schinz, H., 1844, Systematisches Verzeichniss a l l e r bis jetzt bekannten Saugethiere, oder Synopsis Mammalium, nach dem Cuvierschen System, v o l . 1., Solothurn, Jent und Gassmann. Schlosser, M., 1887-1890, 1923, Die Assen, Lemur.en, Chiropteren, Insectivoren, Marsupialier, Creodonten und Caraivoren des Europaischen Tertiars. Beitr. Palaont. Geol. Osterreich-Ungams 6, 7, 8: 1-492. Grundzuge der Palaontologie (Palaozoologie) von Karl A. Z i t t e l , II abteilung - Vertebrata. Neuarbeitet yon'F. B r o i l i und M. Schlosser. Munich und Berlin. Oldenbourg, pp. V plus 706 i l l u s . Schmidt, F., 1943. Naturegeschichte des Baum - und des Steinmarders, mit vergleichenden Betrachtungen ihrer nachsten Verwandten, besonders des sibirischen Zobels und des amerikanischen Fichtenmarders. Monogr. Wild-saugetiere, Leipzig, 10: 1-528, translated by A. C. Marshall. Scopoli, G. A., 1777, Introductio ad historiam naturalem sistehs genera lapidum, plantarum et animalium hactenus delecta, caracteribus essentialibus donata, in tribus divisa, subinde ad leges naturae. Prague, (publisher and pagination unknown). --74--LITERATURE CITED continued Scott, W. B.f 1937, A history of land mammals in the western hemisphere. 2nd. edit., New York, Macmillan, pp. XIV plus 786 i l l u s . Sears, P. B., 1935a 1935b, 1938, Glacial and postglacial vegetation. Bot. Rev. 1: 37-51. . Types of North American pollen profiles. Ecol. 16: 488-499. Climatic interpretation of postglacial pollen deposits in North America. B u l l . Amer. Meteoral. Soc. 19: 177-181. 1942, Postglacial migration of five forest genera. Amer. Journ. Botany 29: 684-691. 1948, Forest sequence and climatic change in north-eastern North America since early Wisconsin time. Ecol. 29: 326-333. Seton, E. T., 1929, Lives of game animals, vol. 2 pt. 2. New York, Double-day, pp. 369.-746 i l l u s . Shantz, H. L. and R. Zon, 1924, Atlas of American agriculture. Sect. E, Natural Vegeta-tion. U. S. D. A. Bur. Agric. Econom., pp.. 29. -. Shaw, G., 1800, General zoology, or systematic natural history, vol. 1. London (publisher and pagination unknown). Shelford, V. E., 1935, Sere, climax and influent animals, with special reference to the transcontinental coniferous forest of North Amer-ica. Ecol. 16: 375-402. - 75 -LITERATURE CITED continued Simpson, G. G., 1943, 1945, 1947, Mammals and the nature of continents. Amer. Journ. Sc i . 241; 1-31. The principles of classification and a classification of mammals. B u l l . Amer. Mus. Nat. Hist. 85, pp. XVI plus 350. "Holarctic mammalian faunas and continental relation-ships during the Cenozoic. B u l l . Geol. Soc. Amer. 58* 613-683. Simpson, G. G.» and A. Roe., 1939, Quantitative zoology; numerical concepts and methods in the study of Recent and f o s s i l animals. New York, McGraw-Hill, pp. XVIII plus 414 i l l u s . Stirton, R. A., 1V39, Significance of Tertiary mammalian faunas in holarctic correlation with.special reference to the Pliocene in California. Journ. Paleont. 13: 130-137. Swarth, H. S., 1911, Birds and mammals.of the 1909 Alexander Alaska expedi-tion. Univ. C a l i f . Publ. Zool. V : 9-172. 1912, Report on a collection of birds and mammals from Vancouver Island, Univ. C a l i f . Publ. Zool. 10: 1-124. Taylor, G., 1937, Environment, race and migration; fundamentals of human distribution; with special sections on raci a l classification and settlement i n Canada and Australia. Univ. Toronto Press, pp XV plus 483 i l l u s . Thwaites, F. T., 1934, Outline of gl a c i a l geology. Ann Arbor, Edwards, pp. II plus 115 V i l l u s . - 76 -LITERATURE CITED continued Thomas, 0., 1911, The mammals of the tenth edition of Linnaeus; an attempt to f i x the types of the genera and the exact bases and loca l i t i e s of the species. Proc. Zool Soc. London 1911: 120-157 Thomas, 0., and G. S. M i l l e r , 1905, The generic names given by Frisch in 1775. , Ann. Mag. Nat. Hist. ser. 7, 16: 461-465. Transeau, E. N., 1903, On the geographic distribution and ecologic relations of the bog plant societies of northern North America. Bot. Gaz. 36: 401-420. 1905 Forest centers of eastern America. Amer. Nat. 39: 875-889. True, F., 1884, A provisional l i s t of the mammals of North and Central America, and the West Indian Islands. Proc. U. S. Nat. Mus. 7: 587-611. Turton, W. 1802, A general system of nature through the three grand kingdoms of animals, vegetables and minerals; systematically divided into their several classes, orders, genera, species and varieties,,with their habitations, manners, economy, structure and peculiar i t i e s . By Sir Charles Linne.... In 7 volumes, vol. Animal Kingdom. London, Lackington and Allen. Ullman, M. B. (Editor), 1949, S t a t i s t i c a l abstract of the United States 1949. U. S. Dept. Commerce, Bur. Census, pp. XV plus 1122 .".illus. LITERATURE CITED continued Voss, J., 1934, Postglacial migration of forests i n I l l i n o i s , Wisconsin, and Minnesota. Bot. Gaz. 96 i 3-43. 1937, A comparative study of bogs on Cary and Tazewell d r i f t . Ecol. 18: 119-135. 1939, Forests of the Yarmouth and Sangamon interglacial periods in I l l i n o i s . Ecol. 20: 517-528. Wagler, J. G., . 1830, Naturliches System der .. Amphibien, mit vorang Classification der Saugethiere und Vogel. Stuttgart, (publisher and pagination unknown). Wilson, L. R., . 1938, The postglacial history of vegetation in north-western Wisconsin. Rhodora 40: 137-175. Wynne-Edwards, V. C , . 1937, Isolated arctic-alpine floras in eastern North America: a discussion of their g l a c i a l and Recent history. Trans. Royal. Soc. Can. 3: 1 - 26. Yeager, L. E., ; 1941, A contribution toward a bibliography on North American fur animals. 111.. Nat. Hist. Surv. B i o l . Notes 16, Publ. 338, pp. 209. 1950, Implications of some harvest and habitat factors on pine marten management. Irans. 15th N. Amer. Wildl. Cong., pp. 319-334. - 78 -APPENDICES Definitions of measurements used: 1. Condylobasal length: the greatest antero-posterior distance between the anterior end of the premaxillae,and the posterior ends of the condyles. 2. Rostral width: the narrowest width of the rostrum behind the canines. 3. Upper tooth rows the shortest distance separating the anterior surface of the upper canine base from the posterior surface of the hypocone of the upper molar. 4. Canine width: the greatest width separating the upper canines at their bases. 5. Bullae length: the shortest length of the auditory bulla, taken from the medial and posterior surface of the foramen lacerum canal to the midpoint of the suture on the jugular process. 6. Upper molar width: the greatest width of the upper molar, when taken in a late r a l plane. 7. Upper molar length: the greatest length of the inner moiety of the upper molar, taken in an antero-posterior plane,., Characters 4 and 5 were calculated as a ratio (canine width divided by bulla length) APPENDICES continued B. Provisional a r t i f i c i a l key to the skulls of the North American marten. This key has been designed for skulls and uses skull characteristics insofar as specimens from which to devise them have been available. Where these have been lacking, the key reverts to a purely distributional one. Many specimens w i l l not be identifiable using this key. This is to be expected since i t w i l l be usable only i f they are representative of what at present appears typical for that sub-species. A Canine width - / - bulla length nearer .90 than 1.50 in males, and nearer .85 than .95 in females americana group (a) From continental North America, excluding the Kenai: Peninsula, Alaska. 1. Condylobasal length nearer 8.5 than 8.0 cm. in males, and nearer 7.6 than 7.3 in females, a Upper molar length plus width nearer 1.35 than 1.20 cm. in males, and nearer 1.20 than 1.05 cm. i n females.-.^ (1) From east of Hudsons Bay brumalis (2) From west of Hudsons Bay abieticola b Other than above; from west of Hudsons Bay actuosa - 80 -APPENDICES continued 2. Other than above a Canine width - / - bulla length nearer .94 than. .98 in males, and nearer .80 than .84 in females. From east or south of Hudsons Bay ......americana b Other than above; from west of Hudsons Bay • • • .abietinoides (b) From Newfoundland, or the Kenai Peninsula 1. From Newfoundland atrata 2. From the Kenai Peninsula kenaiensis B Other than above caurina group (a) Not from California 1. Rostral width nearer 1.90 than 1.65 cm. in males, and nearer 1.60 than 1.45 in females. From Queen Charlotte Islands or islands to the north.. .nesophila 2. Other than above a. From the mainland (1) From the Coast range north of California caurina , (2) From the Rocky Mountains south of the 49th parallel ....... .origenes b. From Vancouver Island vancouver-ensis. (b) From California or extreme western Nevada 1. From the Coast Range humboldtensis 2. From the Sierra Nevadas sierrae