right ramus, C. I. T. No. 3943, and is complete except for the canine. The other two bear no numbers and I have designated them "A" and "B." "A" is a left ramus with the dentary broken off anterior to the canine and bears p4 and the canine. "B" is a right ramus bearing m2 and the roots of m3 and is broken off at the middle of the alveolus of m1. Each specimen has certain peculiarities but they resemble one another so closely that I regard all three as of the same species. The teeth, where comparable, are of essentially the same size and configuration. The horizontal rami of the dentaries are the same. The fossils differ, however, in the configuration of the coronoid process. In No. 3943 the coronoid is robust and inclined anteriorly with respect to a line drawn perpendicular to the dentary. The posterointernal ramal fossa (see Hibbard, 1953) is deeply excavated with a distinct superior border approximately halfway between its inferior border and the top of the coronoid. In addition to the mandibular foramen there is a small foramen immediately posterior to it opening into the posterointernal ramal fossa. I shall refer to this as the post-mandibular foramen. The tip of the coronoid is broad, not tapering, and quadrate, and its entire superior border is inclined rather sharply medially. Specimen "B" differs from No. 3943 in that the posterointernal ramal fossa extends nearly to the tip of the coronoid, which is narrower toward the tip and somewhat tapered dorsally. The post-mandibular foramen is large and the opening of the mandibular canal is within the posterointernal ramal fossa. In addition the coronoid does not incline anteriorly. Specimen "A" is intermediate between No. 3943 and "B" in the characters mentioned and differs from both in that the post-mandibular foramen is widely separated from the mandibular foramen.

Comparison of the Pleistocene specimens with specimens of Recent species of North American Sorex reveals that the presence or absence of the post-mandibular foramen is almost constant in any one species. In possessing this foramen the fossils differ from most individuals of the species: Sorex cinereus, S. pacificus, S. milleri, S. vagrans, S. obscurus, S. ornatus, S. fumeus, S. palustris, S. bendirii, and S. veraepacis. The fossils differ from all these species in other characters as well; consequently detailed comparisons with them need not be made here. Species which possess the post-mandibular foramen include Sorex saussurei, S. merriami, S. trowbridgii, S. arcticus, S. tundrensis, and S. sclateri. Sorex merriami differs in smaller size, smaller and weaker dentition, relatively higher coronoid, and relatively deeper and shorter dentary. Sorex trowbridgii is similar to the fossils and to S. saussurei. Differences between the jaws of S. trowbridgii and S. saussurei seem to me to be differences of size only. Sorex sclateri is larger than the fossils and m2 is longer in relation to m1, being almost the same size as m1. In the fossils m2 is noticeably shorter than m1, owing to close appression of the hypoconid and protoconid and in general to a smaller talonid area. Sorex arcticus differs in larger incisor and p4. Sorex tundrensis differs in relatively narrower molars. I have compared the fossils also with the Pliocene and Pleistocene Sorex taylori Hibbard, and find that the fossils are larger and have larger teeth and a much wider separation of the protoconid and metaconid. I can find no significant way in which the fossils differ from S. saussurei. This of course implies similarity to S. trowbridgii. Since S. saussurei is a widespread species in México today and since it occurs in the vicinity of the San Josecito area the specimens under discussion are referred to this species.

Cryptotis mexicana