of the past. A dent began to form on one side of 

 the round egg; the first definite sign that it had 

 ceased really to be an egg; it was now in the gas- 

 trula stage, a beginning larva. As the dent grew 

 deeper, the invagination thus formed finally con- 

 nected with the hollow interior, making what was 

 later to be the mouth to the primitive digestive 

 cavity. The outer surface of the gastrula now be- 

 came covered with cilia, and the larva escaping 

 from its egg-shell swam to the surface of the water. 



Three days later, a look at my larvae showed that 

 they had elongated considerably; when seen from 

 the side they were roughly triangular; when 

 viewed from above they were somewhat boat- 

 shaped but blunt at the ends. Underneath, they 

 carried a keel, not in the usual fore and aft direc- 

 tion, but transversely, from beam to beam. This 

 keel or ridge was fixed well aft, while immediately 

 following it was the mouth, now apparently trav- 

 eling toward the stern. Here it was that I first 

 noticed that a spicular skeleton was starting to 

 form. 



It was the forming of this skeleton that decided 

 me to experiment and try for an answer to the 

 problem of their unique distinction among the 



[101] 



