from Indiana, gives a basis for applying the name Synaptomys cooperi gossii to the specimen from Knoxville. This is the same name recently used by Fichter and Hansen (Bull. Univ. Nebraska State Mus., 3(2):2, September, 1947) for the Iowan specimens, although they seemingly applied the name without being aware of Bole and Moulthrop's earlier naming of S. c. saturatus (Sci. Publs. Cleveland Mus. Nat. Hist., 5:149, September 11, 1942).

Howell (N. Amer. Fauna, 50:30-31, August 5, 1927) had only eight specimens of this subspecies available when he revised the genus Synaptomys. Of these eight (Maine: Mount Katahdin, 2; New Brunswick: Near Bathurst, 1; New Hampshire: Fabyans, 1, the type; Quebec: St. Rose, 4), only the type and one of the specimens from St. Rose are adults. Concerning the others, Howell wrote (op. cit.:31): "The example from near Bathurst is not adult and has a damaged skull, so is identified provisionally. All other specimens are too young for positive diagnosis."

Since Howell's revision only one additional specimen has been reported. Anderson (Ann. Rept. Provancher Soc. for 1939, p. 71, 1940) reported it from Table Mountain, 3888 ft., Gaspé County, Quebec.

In the collection of the University of Kansas Museum of Natural History there is still another specimen. It is an adult male topotype (No. 6483 KU, formerly No. 72 in the collection of Alfred E. Preble) obtained on August 21, 1905, at Fabyans, New Hampshire. The measurements of this specimen are as follows (measurements in parenthesis are those of the type as given by Howell, op. cit.): Total length, 135 (132); tail, 26 (24); hind foot, 22 (20); condylobasilar length, 25.1 (25.8); rostral length, 6.5 (6.8); rostral breadth, 4.7 (4.9); interorbital breadth, 3.3 (2.8); zygomatic breadth, 15.4 (16.0); lambdoidal breadth, 12.1 (12.4); incisive foramina, 5.9 (5.7); height of skull, 9.1 (9.3).

Howell (op. cit.:30) characterized S. b. sphagnicola as: "Large and high [skull] with narrow interorbital sharply ridged, the ridges of the type being joined for a distance of 4 millimeters; interparietal narrow and rectangular. The rostrum is long, tapering very little, and the nasals, slightly constricted medially are quite narrow posteriorly. The incisive foramina are long and wide." Howell further stated (op. cit.:30-31) that: "It is hard to predict what will be found to constitute the most valuable cranial characters in distinguishing this race from adult medioximus. The discernible differences now are in the shape of the interparietals, rostral characters, and interorbital differences that will probably not hold good when animals of the same age are compared."

As can be seen from a comparison of the measurements given above for the type and the topotype, some of the characteristics given by Howell are not found in the topotype: The interorbital region is not narrow (in fact it is wider than it ordinarily is in some other subspecies of Synaptomys borealis) and the incisive foramina are not longer than in other subspecies of Synaptomys borealis.

As far as present material permits us to judge, Synaptomys borealis sphagnicola is characterized, cranially, by: Skull large; interorbital region sharply ridged (the ridges being joined for a distance of 4 mm. in the type and of 4.5 mm. in the topotype); rostrum long, tapering relatively little; nasals slightly constricted medially and unusually narrow posteriorly; interparietal narrow and rectangular.

Prior to 1935 caurinus was considered to be a monotypic species. In 1935 Racey and Cowan (Rept. British Columbia Provincial Museum for 1935, pp. H 25-H 26) examined material from southwestern British Columbia of C. caurinus, including a series of 24 specimens from Alta Lake, and compared it with Clethrionomys gapperi occidentalis and C. g. saturatus. They found caurinus to be distinct from C. g. saturatus but were "not convinced that occidentalis and caurinus both merit systematic recognition; should they prove to be indistinguishable, as the little available material indicates, occidentalis will take precedence on grounds of priority. It is our opinion that further study of the distribution of the genus in British Columbia will lead to the recognition of occidentalis as the form inhabiting coast-line and saturatus the interior of British Columbia" p. H 26. In the face of these opinions Racey and Cowan nevertheless recognized caurinus under the name Clethrionomys gapperi caurinus (Bailey).

In spite of the treatment by Racey and Cowan (op. cit.) of occidentalis and caurinus as subspecies of C. gapperi, later authors arranged occidentalis as a member of the "californicus" group although they retained caurinus in the gapperi group. For example, Davis (The Recent Mammals of Idaho, The Caxton Printers, pp. 307-308, 1939) assigned C. caurinus to the gapperi group, although he regarded C. caurinus as a species (not a subspecies). He regarded also C. occidentalis as a species (not a subspecies) but assigned it to the californicus group. Dalquest (Univ. Kansas Publ., Mus. Nat. Hist., 2:344, April 9, 1948) considered occidentalis to be conspecific with Clethrionomys californicus and wrote (op. cit.:101): "The californicus group, I feel, contains only the races of Clethrionomys californicus, while the gapperi group contains C. gapperi and its races, including caurinus, and possible other species." Dalquest gave no indication that he had examined any specimens of caurinus.

When Dalquest (op. cit.:344) arranged occidentalis and californicus as subspecies of the same species, he used the name combination Clethrionomys californicus occidentalis because he ignored, or was unaware of, the page priority of occidentalis over californicus. We regard the anterior position of occidentalis as nomenclatural priority and therefore employ occidentalis rather than californicus as the specific name.

Differences between the gapperi group and the occidentalis group include: postpalatal bridge (complete in both groups) truncate posteriorly in the gapperi group and with a median, posteriorly directed, spine in the occidentalis group (this character is not evident in all specimens; some gapperi have a spine, and some occidentalis have the spine much reduced); dentition of the occidentalis group is heavier; enamel pattern of M3 and m1 in occidentalis more simplified—the number of salient and re-entrant angles tends to be reduced in adults of the occidentalis group.

An examination of specimens of caurinus (British Columbia: Mt. Seymour, 2 KU; Lund, Malaspina Inlet, 2 USBS; and Inverness, mouth Skeena River, 1 USBS), reveals that, in the presence of the median postpalatal spine and in the characters of the molars, caurinus agrees with the occidentalis group.