BLASTOGENIC VARIATIONS. 



Ill 



the hybrids obtained in May, June, and July were of the 

 almost pure Sphccrechinus type, of which an example 

 is given in Fig. 20; but about a third of them or less 

 were of the intermediate or Strongylocentrotus type, of 

 which an example is given in Fig. 21. In November, 

 on the other hand, only about a sixth of the hybrid 

 larvae were of the Sphcurechinus or maternal type, and 



FIG. 20. Sphcerechinus 

 larva. 



FIG. 21. Sphcereckinus 9 Stron- 

 gylocentrotus $ larva. 



five-sixths of the paternal type. Finally, in December 

 and January, all the larvae were of the paternal type. 

 These so-called paternal larvae in almost all cases 

 showed obvious traces of their hybrid origin, but they 

 were evidently much more inclined to the Strongylocen- 

 trotus than to the SphccrecJiinus type. 



Combining this series of observations wi'th that just 

 recorded, we therefore find that in the summer months, 

 when the Strongylocentrotus sperm is in a condition of 

 minimum maturity, the Splicer echinus 9 Strongylo- 

 centrotus $ hybrids are chiefly of the Splicer echinus 

 type. As, however, the maturity of the sperm in- 

 creases, it is able to transform first a portion and then 



