AUTOTOMY OF THE TAIL IN RODENTS. 5 



The writers offer no speculations as to the physico-chemical 

 factors which bring about this excessive growth of the hair on 

 the severed stump of the tail of Perognathus. A superficial 

 analogy suggests itself between this terminal tuft and the cluster 

 of adventitious shoots which arise from the stump of a tree. 

 Possibly the analogy is more than superficial. 



How general this process of autotomy is among the pocket- 

 mice we can not say. We have noted its occurrence in only 

 one other species, P. panamintinus bangsi, a small desert form 

 in which the process is closely similar to that in P. fallax. 



To what extent this phenomenon may be spoken of as a 

 "protective adaptation," and whether or not it arose through 

 natural selection, we are unwilling to conjecture. It is more 

 than possible that a pursuing carnivore would sometimes be 

 cheated of its prey in this manner, since the greatly elongated 

 tail might readily be seized by the claws or teeth. The asso- 

 ciation between the fragile structure of this appendage and the 

 peculiar instinctive responses of the animal is probably not 

 accidental. 



A phenomenon similar to that discussed above, but differing 

 in essential features, has been observed by one of us in a species 

 of Peromyscus (P. boylei rowleyi). This mouse likewise has a 

 relatively long tail, and undergoes, when this member is seized, 

 very much the same curious gyrations as does Perognathus. In 

 the former species, however, the vertebrae are not broken, nor 

 indeed is the central axis of the tail severed at all. The skin 

 breaks at some point in its length, and slips off, leaving in one's 

 hands the long tubular sheath which covered the appendage. 

 This process is of such frequent occurrence as to be quite charac- 

 teristic of these mice, which thus contrast strongly with the 

 various other species of Peromyscus that we have observed. In 

 the subspecies of P. maniculatus , for example, w r e have never 

 observed the detachment of any portion of the tail or of the 

 skin, in the course of handling several thousands of these animals. 

 Nor have we observed it in any other genus of California Muridas. 



In Peromyscus boylei rowleyi we must again note what seems 

 to be a significant correlation between structure and behavior. 

 The skin of the tail breaks and slips off with remarkable facility. 



