TROPHOBLAST AND SEROSA. 597 



becoraes reflected over the surface of the yolk, while the serosa 

 shrinks away towards the anterior end of the egg (Fig. 3, b). 

 Finally the serosa shrivels up until it constitutes the dorsal 

 organ of the insect embryo, which in primitive in- 

 sects is situated in the nuchal region, e.g. Lepisma, 

 Gryllus, Forficula (cf. Fig. 4). The dorsal organ is subse- 

 quently withdrawn into the yolk, where it undergoes disinte- 

 gration and absorption. 



In the more primitive insects, therefore, the point at which 

 the absorption of the serosa into the yolk takes place, lies in 

 the nuchal region. 



The resemblance between the trophic vesicle of the embryo 

 ofPeripatus novse-britannise shown in Fig. 3, a, with the 

 serosa of the embryo of Gryllus copied from Heymons in Fig. 

 3, B, is too striking to need further comment. If this were all, 

 I should almost regard the suggested homology between the 

 trophoblast of my Peripatus embryos and the serosa of insect 

 embryos as a fait accompli. But it is not all. 



The indusium of the Locustidse described by Wheeler (12), 

 and the dorsal organ ('' micropyle," " micropylar organ ^') of 

 the Poduridse have to be taken into consideration. They can 

 only be accounted for by invoking the aid of hypothesis ; but 

 then every explanation of any structure or phenomenon is 

 more or less hypothetical. 



Since the embryos of Peripatus novse-britannise have 

 enabled us for the first time, so far as the Invertebrata are 

 concerned, to seize upon the most precious conception of the 

 trophoblast, I think we are not only justified but absolutely 

 called upon to face any difficulties which may obscure the 

 application of this conception in all its phylogenetic ramifica- 

 tions. 



To account for the dorsal organ of the Poduridae and the 

 indusium of the Locustidffi, we must call to our aid the prin- 

 ciple of substitution. Without doubt Kleinenberg has left us 

 a noble legacy in his principle of substitution, applicable as it 

 is both to ontogenetic and phylogenetic changes. 



I assume that the transformation of the trophoblast into the 



