The San Josecito collection contains 22 rami of a species of _Cryptotis_. Many are nearly complete although none possesses the incisor. In addition there is a rostrum that on the right side bears the last two unicuspids, P4, M1, and M2. I have compared these fossils with specimens of the following species of _Cryptotis_: _C. mexicana_, _C. magna_, _C. nelsoni_, _C. thomasi_, _C. alticola_, _C. parva_, _C. orophila_, _C. pergracilis_, _C. guerrerensis_, _C. obscura_, _C. mera_, _C. soricina_, _C. fossor_, _C. goodwini_, _C. griseoventris_, _C. meridensis_, _C. mayensis_, and _C. micrura_. The four species first mentioned and the fossils seem to fall into one group. The remaining species fall into another group characterized by a smaller occlusal area of the talonid on all three molar teeth with respect to the trigonid, and especially by the smaller and weaker talonid of m3 which possesses only one bladelike cusp, the hypoconid. In the first four species the talonids are larger than in the other species when compared to the trigonids, and the talonid of m3 possesses a well developed hypoconid and entoconid with a distinct basin between them. The rami of San Josecito specimens closely resemble those of _C. mexicana_ in both size and qualitative characters. The rostrum mentioned above differs from those of _C. mexicana_ in that the unicuspids are larger, especially the posteriormost one. _Cryptotis thomasi_ and _C. magna_ are eliminated from consideration here on geographical grounds. Little difference may be seen between the rami of _C. mexicana mexicana_ from Veracruz and _C. nelsoni_. The fossils are referred to the former species since it has a rather wide distribution in México in contrast to _C. nelsoni_ which is restricted to Volcán de Tuxtla, Veracruz. The northernmost Recent occurrence of _C. mexicana_ known to me is from Las Vigas, Veracruz. As far as I know the species has not previously been recorded from the Pleistocene. Stirton (1930:225), in summarizing the group characters of Recent Soricidae, listed the bicuspid talonid of m3 as one of the characters of the "_Blarina_ group," which includes _Neomys_, _Soriculus_, _Notiosorex_, _Chimarrogale_, _Cryptotis_, and _Blarina_. That this character does not obtain universally within the group is demonstrated by the unicuspid structure of the talonid of m3 of the majority of the species of Mexican _Cryptotis_. I am grateful to Drs. David E. Johnson and Henry W. Setzer of the United States National Museum, and to Dr. John Aldrich and Miss Viola S. Schantz of the United States Fish and Wildlife Service for permitting me to examine specimens in their care. Also I am obliged to Professor E. Raymond Hall for permission to study the specimens from San Josecito Cave. The late Professor Chester Stock intrusted the specimens to Professor Hall for study and description. LITERATURE CITED CUSHING, J. E., JR.