sufficiently cleared that the stained baculum could be seen easily, the solution was replaced by glycerin in which clearing was completed. The time required for the entire process varied from one day to more than two weeks depending on the size of the specimen and on its condition. Fresh specimens clear more rapidly than dried specimens, and those that are dried more rapidly than those that are preserved. A three or four per cent solution of hydroxide will hasten the process, but more frequent observation is required to prevent excessive maceration.

Specimens were then examined in a shallow dish containing glycerin under a binocular microscope. The baculum can be viewed from any desired direction. The method described above leaves the baculum intact within the glans penis; therefore its orientation can be determined relative to the thick walled urethra and the thin walled dorsal artery that extends onto the dorsal side of the baculum. The ventral curvature of the penis proximal to the baculum, and the distal extension, characteristic of most species, of the dorsal border of the glans (both shown in Figure 1) are other features aiding in correctly orienting cleared specimens. The digitate processes are not so often injured, lost, or displaced when the method described above is used as they are when the penis is dissected. Specimens were stored in glycerin in glass shell vials having polyethylene stoppers. A small card bearing the name, number, locality, and other data was placed in each vial. A specimen thus enclosed can be kept indefinitely, or removed and mounted in balsam as described by White (1951:631) or in plastic as described by Dearden (1958:541) and thus stored in the vial containing the skull of the specimen.

Drawings were made on millimeter ruled paper while the baculum was viewed under a binocular microscope with a square ruled eyepiece.

Unless otherwise noted all specimens listed are in the University of Kansas Museum of Natural History. Catalogue numbers are cited. Measurements are accurate to within less than one-tenth of a millimeter. Proportions as stated in the text are approximations, accurate to within one-twelfth (8.33 per cent). The range of variation is unknown for some species. Mention is made if maturity is known or suspected to differ in specimens being compared.

The development of the baculum has been studied by Callery (1951) in Mesocricetus auratus and by Ruth (1934) in the laboratory rat. In the rat (Rattus norvegicus) the bone is of endoblastemal origin being laid down by a condensation of undifferentiated mesenchymal cells. At the distal end of the bone dense fibrous tissue is then differentiated and at the proximal end hyaline cartilage. Growth is by substitution at the proximal end and by subperiosteal lamellation circumferentially. A marrow cavity is formed by resorption. In the baculum of the hamster the primary center of ossification is in the stalk, and is present at the age of three days; the secondary centers are in lateral processes and are present at 80 days and enlarge subsequently. A tertiary center, in each median process, may or may not develop later. Maximum development of the baculum is reached late in the reproductive life of the hamster.

The early ossification of the baculum noted in the rat and the hamster occurs in Microtus also. A specimen of Microtus montanus fusus (76831, from 5 mi. N, 26 mi. W Saguache, 9600 ft., Saguache County, Colorado) only 74 mm. in total length and weighing only 6.6 grams, had a slender ossified baculum having enlarged ends. This vole was one-half of the average length and less than one-fifth of the average weight of an adult, and of approximately the size at which weaning takes place.

The development of the baculum in Microtus ochrogaster was studied in 32 specimens of various ages. The specimens (between Nos. 74994 and 75074) were collected between August 15 and September 4, 1957, at localities on the Great Plains. These specimens were from breeding populations, as evidenced by pregnancy of females and by large size of testes of males. The length and width of the stalk of the baculum, the presence of digital ossifications, the total length of the animal, and the size of the testes were noted. Variability in length of testes is greatest when voles are from 140 to 150 mm. in total length. Sexual maturity is reached rather abruptly when the total length of most individuals is 140 to 150 millimeters. If the baculum likewise underwent more rapid growth at the onset of sexual maturity, greater variability should be evident in the length of the baculum of voles 140 to 150 mm. in total length than in bacula of voles of other sizes. This was the case (see Figure 1d). The baculum does not, however, suddenly reach its maximum maturity.

The primary ossification is in the stalk. The secondary ossification is in the median process except in Lagurus (Dearden, 1958:551) and some individuals of Neofiber (see account on page 258). Tertiary centers of ossification are in the lateral processes. The primary ossification is present at an early age and subsequently increases in size and solidity. The secondary and tertiary ossifications are progressively more common in older voles. The increase in degree of ossification of all parts continues after sexual maturity is reached. Individual variation and variation with age in the baculum of Microtus pennsylvanicus have been illustrated by Hamilton (1946:380). Figures 14, 15, and 17 illustrate variation with size, which is correlated with age, and also illustrate individual variation. The three bacula are from adult voles having testes that measured 15, 16 and 16 mm. in length, respectively. Each vole was trapped in late June. The total lengths in millimeters of the three voles are 172, 167, and 181; weights are 55, 52.4, and 65.5 grams. I judge that the greater size of the stalk and the better developed base shown in Figure 17 than in Figure 15 are illustrative of age variation; the difference in the size of the lateral digitate processes is, in this case, attributable to individual variation. Differences in the distal end of the baculum in Figures 42 and 43, show individual variation also. Figures 35 and 36 represent two different subspecies; different individuals of M. mexicanus mogollonensis, however, exhibit individual variation of the same degree.

Hall and Cockrum (1953) list 44 species of microtines in North America. At least twelve of these are insular or local forms perhaps derived from some other species; for example Microtus coronarius, an insular form derived from Microtus longicaudus; Microtus provectus, considered by Chamberlain (1954:587) and by Wheeler (1956:176) as a subspecies of Microtus pennsylvanicus; and Microtus ludovicianus, a close relative of Microtus ochrogaster.

All North American genera have been studied. Of the genus Microtus in North America, all subgenera but Orthriomys and all species but the following nine, have been studied: M. (Orthriomys) umbrosus, the insular M. (Stenocranius) abbreviatus, M. (Microtus) breweri, M. (Microtus) nesophilus, M.

Figures 2-13. Bacula of microtines. Unless indicated otherwise views are (a) of the dorsum, (b) the right side, and (c) the proximal end with the dorsal surface upward. Exact localities are given in accounts of species concerned.

1. Lemmus trimucronatus, 50678, Point Barrow, Alaska.
2. Dicrostonyx groenlandicus, 50539, Porcupine Lake, Brooks Range, Alaska.
3. Dicrostonyx