If, after the postjuvenal molt, a distinct adult pelage is acquired it is difficult to separate it from the annual replacement of pelage in adults at the beginning of the rainy season. Adults of both species have been found in molt in all months of the year. To the north, in Texas, the pelage of winter-taken specimens is denser and slightly more reddish than that of specimens taken in spring and summer. In the two last mentioned seasons, the pelage is more uniformly gray. To the south, in México, the pelage is heavy and long in most specimens taken in the rainy season. The percentage of specimens in molt immediately before the rainy season and immediately before the dry season is slightly higher than in specimens taken at other times of the year. The adult or seasonal molt (both loss of old pelage and growth of new) resembles that in Peromyscus truei gilberti, described by Hoffmeister (1951:6) as proceeding "posteriorly as a wave over the entire back." The new hair is slightly brighter than the old. Old adults are usually in ragged pelage regardless of season; possibly only one regular annual change of pelage occurs in most animals before they die. Only one case of melanism was observed among all the specimens of both species examined. It was a young male B. t. taylori, KU 35943, from 6 mi. SW San Gerónimo, Coahuila, possessing black hairs throughout. Its hairs are longer and finer than those on specimens of comparable age and sex. No albino was found, although Stickel and Stickel (1949:145) record one—an adult male of B. taylori.
TAXONOMIC CHARACTERS AND RELATIONSHIPS
External parts.—Length of body, foot, ear, and tail are useful when considered together in distinguishing species and subspecies. I found as Hooper (1952a:91) did that length of ear in combination with length of hind foot suffices to identify nearly all specimens to species, especially where the two species occur together.
Pelage.—Color in adults is of especial value in subspecific determination; the manner in which it varies geographically is described on pages 609, 630.
Skull.—Difference in occipitonasal length and zygomatic breadth, both having low coefficients of variation, are useful in separating species, especially where they are sympatric. Shape of presphenoid, nasals, interparietal, frontoparietal sutures, and length and degree of the openings of the incisive foramina are useful in delimiting subspecies. The rostrum of B. taylori, in front of the frontonasal suture, is deflected three to five degrees ventrally in 85 per cent of the adults examined, and in B. musculus is less, or not at all, deflected.
Teeth.—Alveolar length of the upper and lower molar tooth-rows aids in distinguishing fossil and Recent species, and to a lesser degree in delimiting subspecies. Occlusal pattern is useful in estimating the relationship of fossil and living species. Degree of development of the mesostyle, mesostylid, mesoloph, and mesolophid have been useful in determining relationship between fossil and living species as well as useful in separating the living species. Rinker (1954:119) and Hooper (1957:48) have shown the degree of variation in dental patterns in Peromyscus, Sigmodon, and Oryzomys, mice thought to be closely related to Baiomys. In pygmy mice, however, the dental patterns are relatively constant. The lophs and styles are subject to some geographic variation but, nevertheless, are useful in estimating relationships.
Fig. 5. Ventral view of hyoid bones. × 18.
A. Baiomys musculus brunneus, adult, female, No. 30182 KU, Potrero Viejo, 1700 feet, Veracruz.
B. Baiomys taylori analogous, adult, female, No. 36761 KU, 2 mi. N Ciudad Guzmán, 5000 feet, Jalisco.
Hyoid apparatus.—Shape and, to a lesser extent, size of the hyoid apparatus differentiate nearly all specimens of B. taylori from all those of B. musculus. The hyoid of B. taylori differs from that of B. musculus principally in the shape of the basihyal. It possesses an anteriorly pointed entoglossal process in B. musculus, and is not rounded to completely absent as in B. taylori (see [Figure 5]). The shoulders of the basihyal protrude anteriorly in B. musculus, and are not flattened as in B. taylori. The total length was measured in a sample of 55 basihyals of B. musculus, and was compared to the total length of a sample of 80 basihyals of B. taylori. The means of the two samples differ significantly at the 95 per cent level; the mean plus two standard errors of B. musculus and B. taylori, are, respectively, 2.43 ± .02; 2.18 ± .03. There is sufficient overlap of the samples (mean plus one standard deviation of B. musculus and B. taylori, respectively: 2.43 ± .15; 2.18 ± .15) to make the total length of the basihyal of only secondary importance in distinguishing species, but shape and total length of the basihyal, when considered together, serve to identify all specimens to species. When length of the basihyal is plotted against occipitonasal length (see [Figure 6]), all specimens studied, regardless of age or geographical origin, were separated at the level of species. The hypohyals of B. taylori seemingly remain distinct throughout life; those of B. musculus completely fuse in some adults. The ceratohyals are highly variable in shape and of little taxonomic use.