REMOVAL OF NASAL PITS AMBLT STOMA EMBRYOS 37 



season. Five of these were isolated and their feeding reactions 

 tested with small bits of freshly killed Amblystoma larva. Sixty- 

 three tests were made, fifty-two resulting in positive reactions, 

 a percentage of eighty-three. The food was deposited by means 

 of a capillary pipette some 5 to 6 mm. from the head of the 

 larva. Copeland has pointed out that the reaction of Diemyetylus 

 to food is a complex one involving an approaching reaction as a 

 result of a visual stimulus and a seizing reaction due to the 

 olfactory stimulus. In the present experiments on Amblystoma, 

 it was often evident that the placing of the food attracted the 

 attention of the larva. But the stimulus thus imparted was 

 seldom followed by approach to the source. When it was, the 

 test was excluded from the tabulation. If the larva became 

 quiescent after the visual reaction, it would react positively in 

 the majority of cases to the olfactory stimulus by snapping at 

 the food, as indicated above. 



In addition to the above tests, bits of larvae were placed in 

 the aquarium as far from the individual under observation as 

 possible, so that stimulation of the optic sense could be practi- 

 cally eliminated. In all of the cases, there was no immediate 

 reaction, though the food was eventually discovered and eaten. 



An interesting observation deserves mention in connection 

 with these experiments. Powers in 1907, in studying variations 

 in A. triginum, found that certain of the individuals collected, 

 possessed a remarkable width of head with a correlated expanse 

 of jaw and unusual development of teeth. After watching these 

 individuals, he concluded that the differences ''were due to the 

 cannibahstic" habits of the larvae. They would quickly devour 

 young larvae and snap off projecting parts of their fellows. He 

 found only a few such individuals, and concluded the variation 

 was rare. 



The larvae in the laboratory at New Haven are probably of 

 the species punctatum, though none have as yet been raised 

 through metamorphosis. But all of these, whether reared from 

 eggs in the laboratory or brought in already hatched, would 

 readily eat one another, or such parts of limbs, gills and tail as 

 were obtainable. This cannibalism did not, however, appear in 



