subgenus Eutamias (Eutamias sibiricus asiaticus from Manchuria). Numerous other specimens were examined but not in such detail.

I am grateful to Professor E. Raymond Hall for guidance in the study. For encouragement and advice I am grateful also to Doctors Robert W. Wilson, Cecil G. Lalicker, Edwin C. Galbreath, Keith R. Kelson, E. Lendell Cockrum, Olin L. Webb, and others at the Museum of Natural History, and in the Department of Zoology of the University of Kansas. My wife, Alice M. White, made the drawings and helped me in many other ways. For lending specimens I thank Dr. David H. Johnson of the United States National Museum, and Dr. George C. Rinker of the Department of Anatomy, University of Michigan.

Assistance with field work is acknowledged from the Kansas University Endowment Association, the National Science Foundation, and the United States Navy, Office of Naval Research, through contract No. NR161 791.

Evaluation of Characters

The following paragraphs treat the characters listed by Howell, Ellerman, and Bryant, and such additional characters as I have found useful in characterizing the genera and subgenera of chipmunks. Some of the findings, I think, illustrate how study of such mammalian structures as the baculum, malleus, and hyoid apparatus—structures that seem to be little influenced by the changing external environment—clarifies relationships, if these previously were estimated only from other parts of the anatomy of Recent specimens.

The structural features and characters to be discussed, or listed, below may be arranged in three categories as follows: 1) Characters in which the subgenera Eutamias and Neotamias agree but are different from the genus Tamias; 2) Characters in which the subgenus Eutamias and the genus Tamias agree but are different from the subgenus Neotamias; 3) Structural features that are too weakly expressed to be of taxonomic use.

Characters in which the Subgenera Eutamias and Neotamias Agree, but Differ from the Genus Tamias

Structure of the Malleus.—The malleus in chipmunks is composed of a head and neck, a manubrium which has a spatulate process at the end opposite the head, and a muscular process situated about halfway between the spatulate process and the head of the malleus. An articular facet begins on the manubrium near the neck and spirals halfway around the head of the malleus. A lamina extends from the anterior edge of the head and neck, tapers to a point and joins the tympanic bulla anteriorly where there is a suture between the lamina and bulla. The lamina is one half as long as the rest of the malleus (see figs. 1-3).

The head of the malleus in Tamias is clearly more elongated than in Eutamias. The plane formed by the lamina in Eutamias makes an angle of approximately 90 degrees with the plane formed by the manubrium; in Tamias the two planes make an angle of approximately 60 degrees.

Examination of series of mallei of Eutamias and Tamias indicate that there is slight individual variation, slight variation with age, and no secondary sexual variation. Intraspecific variation in the subgenus Neotamias is slight, consisting of differences in size. Specimens of the subgenus Eutamias from Manchuria have mallei which are morphologically close to the mallei of the subgenus Neotamias.

Figs. 1-3. Dorsomedial views of left malleus.

Fig. 1. Tamias striatus lysteri, No. 11920 sex?; from Carroll Co., New Hampshire.

Fig. 2. Eutamias sibiricus asiaticus, No. 199637 male NM; from I-mien-po, N. Kirin, Manchuria.

Fig. 3. Eutamias townsendii senex, No. 165 male; from Lake Tahoe, California.

Structure of the Baculum.—In discussing the baculum in Eutamias and Tamias, it seems desirable to do so in the light of the structure of the baculum in other sciurids.

The bacula of North American sciurids are divisible into six distinct types represented by those of the genera Spermophilus, Marmota, Sciurus, Tamiasciurus, Eutamias, and Glaucomys.

The type of baculum in Spermophilus is spoonshaped with a ventral process that is spinelike or keellike. Also, spines usually are present along the margin of the “spoon.” The base (proximal end) of the baculum is broad, and some species have a winglike process extending dorsally and partly covering a longitudinal groove. The shaft is more or less curved downward in the middle (see figs. 7, 10).

In Marmota the baculum is greatly enlarged at the posterior end and forms a shieldlike surface. The ventral surface of the base is flattened and the ventral surface of the shaft curves slightly ventrally then dorsally to the tip. The dorsal region of the base culminates in a point, from which there is a ridge that extends anteriorly and that tapers rapidly into the shaft near the tip. The tip, dorsally, has a slight depression surrounded by knobs, which are more or less well defined, and which resemble, topographically, the spines described for Spermophilus (see fig. 8).

In Sciurus the baculum is semispoonshaped and asymmetrical. There is a winglike process on one side and a spine, which projects lateroventrally, on the other side of the tip. The base of the baculum is broad but not so broad as in most species of Spermophilus. Extending posteriorly from the region of the tip, at which point a spine projects lateroventrally, there is a ridge, which is often partly ossified and that extends to a point near the base (see fig. 4).

In Tamiasciurus the baculum is absent or vestigial (Layne, 1952:457-459).

In Eutamias the baculum is broad at the base and the shaft tapers distally to the junction of the shaft and tip, or the base is only slightly wider than any part of the shaft. The tip often forms an abrupt angle with the shaft and there is a keel on the dorsal surface of the tip (see figs. 5, 6).

The baculum in Glaucomys is the most distinctive of that of any American sciurid. According to Pocock (1923:243-244), “The baculum [of G. volans] is exceedingly long and slender, slightly sinuous in its proximal third, and inclined slightly upwards distally. The extreme apex is bifid, the lower process being rounded, the upper more pointed. On the left side there is a long crest running from the summit of the upper terminal process and ending abruptly behind the left side about one-third of the distance from the