LIFE HISTORY OF DESMOGNATHUS FUSCA. 27! 



slightest instinct to go directly to the little pool of water which 

 was now arranged for their reception but would often leave the 

 nest in exactly the opposite direction. In general they made 

 their way, as has been noted above, deeper and deeper into the 

 loose soil until they finally reached the little areas of very shallow 

 water on the surface of the hard soil at the bottom. Under 

 natural conditions the burrows which penetrate the soil in 

 which the Desmognathus lives must furnish natural and easy 

 channels for the descent of the larvae to a level at which the 

 burrows would contain little pools of water. From here it is 

 an easy matter for them to make their way beneath the loose 

 layer of vegetable debris to the open water of the neighboring 

 brook. Possibly an instinct of the larvae to remain in contact 

 with the body of the mother serves as an additional guide to 

 them, as they would thus naturally follow her movements through 

 the burrows. That the larvae, like the adults, are thigmotropic, 

 would also lead them to confine their movements to the crevices 

 in the soil ; a negative heliotropism may also assist in determining 

 the downward direction of their movements. 



On October 5, five more specimens were killed (stage D); of 

 these four were found to have been more or less mutilated, having 

 suffered the loss of a tail, or one of the limbs or both tail and 

 limbs. Since it frequently happens that a considerable pro- 

 portion of larvae collected from their natural habitat show 

 mutilations similar to these, it is evident that the cause in 

 question is one which is operative under natural conditions. It 

 is possible that the injuries are inflicted by the stronger larvae 

 upon their weaker brothers, especially as the injured individuals 

 are usually below the average size. It is more probable, how- 

 ever, that the mother, under the pressure of a paucity of food, 

 found her offspring tempting morsels. The fact that two of the 

 brood were shown by the final count to be missing suggests that 

 this conjecture is the correct one, especially as the larvae them- 

 selves are very hardy and will live for weeks in the laboratory 

 under very adverse conditions. I have noted in several instances 

 similar total disappearance of larvae from terraria in which they 

 had been placed in company with one or two adults, and in two 

 cases have found whole larvae in the stomachs of adults which 

 were in the terrarium from which larvae had disappeared. 



