110 BLASTOGENIC VARIATIONS. 



28 units. From this minimum they then rose rapidly, 

 so that in September they were about 32 units, in Octo- 

 ber and November 34 units, and in December and Janu- 

 ary 35 units in length. The extreme variations were 

 from 36.80 to 24.49, the larvae of this latter length, ob- 

 tained on July 9, being no less than 33.4 per cent, 

 smaller than the former. 



This extraordinary seasonal variation in the size of 

 the larvae is probably very closely, if not entirely, de- 

 pendent on changes in the maturity and nutritional con- 

 dition of the sexual products. Thus, of the specimens of 

 Strongylocentrotus obtained in the winter months, al- 

 most every individual contained ripe sexual products in 

 large quantities, whilst of those obtained in the sum- 

 mer months, not more than about one in four yielded 

 any ova or sperm at all on shaking the ovaries and 

 testes in water, and occasionally twenty or more indi- 

 viduals were opened before any ripe sperm was ob- 

 tained. Again, the best of the specimens obtained in 

 the summer months did not contain nearly so much of 

 the ripe sexual products as they did in the winter. 



That this effect of season on the condition of the sex- 

 cells is more far-reaching than is implied in a mere 

 diminution of size in the resulting offspring, is proved 

 by some observations on the crossing of this species of 

 sea-urchin with another species, viz., Splicer echinus 

 granularis. Hybrids between Splicer -echinus ova and 

 Strongylocentrotus sperm can probably be obtained, 

 though it may be only after several attempts, at all 

 times of the year. It was found, however, that their 

 structure was by no means constant.* The majority of 



*Phil. Trans. 1898, B. p. 470. 



