382 University of California Puhlications in Zoologtj [\'ol. 21 



fig. 1), and of Sorex (pi. 30, fig. 16). Rather comprehensive changes 

 in either of these types are required to accomplish the adult condition. 

 As Doran (1878, p. 444) pointed out, the shrews, in common Avith other 

 insectivores and marsupials, have the processus gracilis firmly united 

 to the tympanic. Where the tympanic lies on the outside of the pro- 

 cessus gracilis in the adult, but on the inside of the processus gracilis 

 in the embryo, the condition seen in adults of the above mentioned 

 genera would seemingly have to be arrived at : by an absorption of some 

 of the parts of the elements that intervene between the processus 

 gracilis portion of the malleus and the area encircled by the tympanic, 

 by a rotation of the elements, or by the formation of the processus 

 gracilis by direct outgrowth from the part of the malleus that is not 

 covered by the portion of the tympanic that spreads over Meckel's 

 cartilage. These same three possible ways in which the adult condition 

 might he brought about present themselves in the case of Oto»permo- 

 phihis. Here the differences in the relative position of the processus 

 gracilis and of the tympanic at embryo and at adult stages are, in the 

 case of Otospermophilus, fully as great as, if not greater than, those in 

 the case of Sarex. In adults of Otospermophilus the processus gracilis 

 is of course free, and directed downward. The material examined does 

 not show which one of the above three general ways of change of posi- 

 tion by the tympanic and processus gracilis was employed, but it does 

 indicate, very strongly, that it was not the last, that is, not outgrowth 

 of the processus gracilis directly from the portion of the malleus that is 

 not covered by a part of the tympanic. From figure 26 it Avill also be 

 noted that the tympanic lies not entirely around the petrous but partly 

 upon the petrous. At this time the tympanic extends over the stapedial 

 process of the incus but terminates at the malleus. The tympanic is 

 much wider posteriorly than anteriorly. This disparity in size between 

 the anterior and posterior parts of the tympanic is rendered more 

 apparent by reason of the fact that the anterior portion, beginning a 

 short distance medially from where it meets Meckel's cartilage, is 

 inclined vertically whereas the posterior portion that extends over the 

 stapedial process of the incus is nearly horizontal. Ossification of the 

 bulla appears to proceed from this ring, no second center of ossification 

 having been noted. In figure 17 of E it will be noted that a constriction 

 is present in the tympanic which divides it into two parts. The basal 

 or inner ring is the smaller of the two. Whether this division into two 

 parts, seen only in E, is a fortuitous one is not known. At any rate 

 it should not be taken to indicate that both a tympanic and an ento- 

 tympanic are present, for earlier stages show no separation into two 



