MODES OF REPRODUCTION 337 



developed the skin is no longer capable of performing respiratory 

 functions, the rapid interchange of gases through it is no longer 

 possible, and therefore this function is transferred to the delicate 

 fin-membranes which bear no scales and which are enlarged for 

 the purpose. It is curious that in this case the required result 

 is not attained by the elongation of the gill-filaments as it is in 

 the embryo of dog-fishes and in the Gymnarchns of the Nile 

 and Gambia. The young of Cymatogaster when born are a 

 little less than \\ in. in length, the maternal fish being from 

 7 in. to a foot. In larger species the new-born young may be 

 from 2 to 3 in. long. They are completely developed in all 

 respects except size, the whole of the development being passed 

 through in the ovary, and at the time of birth the special 

 adaptations of intestine and fin-membranes which have been 

 mentioned have entirely disappeared. As in other cases the 

 structural adaptations of embryonic life last only so long as the 

 conditions which render them necessary. 



The reproduction of the Scorpaenidae has not been so com- 

 pletely investigated. The Scandinavian naturalists merely state 

 that the young of Sebastes norvegicus are born in April or 

 May or even later. The American zoologist Ryder examined 

 a gravid specimen taken off the Banks of Newfoundland in 

 July. He estimated that there were at least a thousand embryos 

 in each ovary, for here the ovaries are not united. These 

 embryos were in an advanced stage of development, though 

 some were still surrounded by the egg membrane, others were 

 free in the ovary. They were slender and only about six 

 millimetres or little more than \ in. long. The yolk-mass 

 was large in proportion to the size of the embyro and con- 

 tained a large yellowish oil-globule. The ovary had very thin 

 inferior and lateral walls somewhat coloured by dark pigment. 

 The dorsal wall was thick and vascular, and from it vascular pro- 

 cesses, divided into slender finger-like branches, dipped down 

 among the mass of embryos. It seems clear therefore that in 

 this case the embyro is nourished by its own yolk, and only de- 

 pends on the maternal tissues for its supply of oxygen which it 

 derives from the vascular ovarian processes just mentioned. 

 The adult Sebastes \s from a foot to two feet in length. Several 

 species of Sebastodes were studied by Eigenmann at San Diego 

 in California. According to his description the eggs are one 



22 



