vol. i.] Hubbard. — Protectivi Devices in Salamanders. 163 



As might be expected, Plethodon has the power to regenerate 

 the tail. Three specimens found in the field had tails aboul a 



centimeter long, white, somewhat translucent and pointed. 

 When autotomy occurs naturally bleeding does not take place. 

 In regenerating, the stump first rounds out with translucent 

 tissue, and then there grows from the middle point a small bud, 

 at first blunt and of uniform diameter afterwards pointed, 

 as shown in fig. 2. It took a month, in an individual whose tail 

 had been amputated, for a tip to grow about four millimeters 

 long, but, as the conditions in this case were artificial, it would 

 not be safe to draw conclusions therefrom as to the rate of regen- 

 eration in nature. 



We may now return to the question of the physiological sig- 

 nificance of the glands. From a comparison of the structure and 

 action of the glands of Plethodon with those of other Batrachia 

 in which the nature of the secretion is known, one is led to sus- 

 pect that the secretion here is poisonous and protective. Numer- 

 ous writers have described tin' product of the various forms of 

 epidermal glands among the Batrachia as milky, acrid, and poi- 

 sonous. Leydig, for example, as one of the older observers in this 

 field, speaks of it as sharp, irritating, benumbing, and capable of 

 producing death. Capparelli has worked out in great detail the 

 poison of Triton cristatus. 



In order to test the action of the secretion I made a number 

 of feeding experiments with all three species here treated, the 

 results of which follow. 



Batrachoseps is eaten greedily, both by the garter snake, 

 Thamnophis elegans, and by the ring-necked snake, Diadophis 

 amabilis. At least five tests were made with the Batrachoseps 

 in connec f ion with these two snakes. The taste is perhaps not 

 quite to the snake's liking, for in some eases there was a slighl 

 gaping after eating, but in no instance was there the least hesita- 

 tion in attack. Only once did regurgitation occur and this once 

 it may have been due to over-eating, for the snake had devoured 

 three or four Batrachoseps in quick succession. 



Diemyctylus, on the contrary, does not seem to be desirable as 

 food. Out of eleven trials, at periods varying from one hour to a 

 day, an individual of Thamnophis elegans only once attacked this 



