4 F. B. SUMNER AND H. H. COLLINS. 



some instances enlarged at the end. This tuft was entirely a 

 new growth, since the middle region of the tail is normally 

 covered with very short hair. As the uninjured tail of this 

 rodent terminates in a "pencil" or tuft of longer hairs (Fig. i) 

 the formation of such a tuft at the point of amputation gave the 

 appearance of a restorative process. There are, however, con- 

 siderable points of difference between the normal "pencil" and 

 the secondary one. The former is produced by the gradual 

 elongation of the hair which covers the tail, commencing with 

 the posterior third or half of this member. The restored tuft 

 (Fig. 2) commences much more abruptly, sometimes being 



FIG. 2. Tail which has undergone autotomy about midway in its length, and 

 upon which a new terminal tuft of hair has been produced. 



confined to the end of the stump. Its hairs are much longer 

 than those of the natural pencil, and their pigmentation is 

 conspicuously decreased, being pale gray on the normally black 

 dorsal side. 



In two instances a small lateral tuft was likewise formed, not 

 far from the end of the stump. In one of these cases, there was 

 found to be an internal injury, corresponding in position to this 

 lateral tuft. 



The capture of two wild specimens with truncated tails indi- 

 cates that this accident may not be uncommon in nature. There 

 was, of course, no clue to the source of the injury. Both speci- 

 mens had developed terminal tufts, one of these being unusually 

 perfect. 



The time required for the formation of a new "pencil" has 

 not been definitely noted. We have record of its having arisen 

 within two months, but we also know that more than four 

 months -may be necessary. 



