34 HAROLD SAXTON BURR 



In the experimental study of this problem, sixteen normal 

 and twenty-four operated larvae were used. Sixteen of the 

 latter were noseless and eight were eyeless. These last were 

 kindly loaned by Dr. Henry Laurens who performed the opera- 

 tions. As soon as the operated individuals as well as the nor- 

 mal ones isolated at the time of the operation began to feed 

 regularly, experimentation was begun. At this time the larvae 

 are about 1 cm. in length. The yolk has completely disappeared, 

 the gills are plume-like arching forward over the head, the fore- 

 limbs are tridigitate, the rudiments of the fourth digit just 

 beginning to appear, and the hind limbs are just noticeable as 

 a slight elevation on either side of the cloaca. Experimentation 

 with these young larvae was difficult because of the verj^ great 

 activity which they exhibited whenever there was any disturb- 

 ance in the water. For that reason the following report deals 

 with somewhat older larvae, the age varying from one and a 

 half to six months. 



The first problem was to determine the relative importance in 

 the obtaining of food of the visual and olfactory sense. For this 

 purpose grains of sand were used, dropped from a capillary 

 pipette so as to fall within striking distance of the larva. A 

 single individual was tested at a time. Four noseless and four 

 normal larvae were subjected to these tests. The first were 

 made when the larvae were five months old, dating from the 

 operation. Another set of exactly similar tests were performed 

 one month. later. 



The forty tests on the four normal larvae resulted in twenty 

 reactions in which the sand grain was snapped at and engulfed, 

 such reactions being designated as positive, and twenty in which 

 no attention at all was paid to the sand. The percentage of 

 positive reactions then, was fifty. Fifty-four tests on four nose- 

 less larvae resulted in thirty-nine positive reactions, a percentage 

 of seventy- two. The same eight larvae tested one month later 

 gave for the normal larvae eleven positive reactions out of forty 

 tests, or 28 per cent and for the operated twenty-two out of 

 forty-four, or 50 per cent (table 1). 



