ORGAN-FORMING SUBSTANCES IN EGGS OF ASCIDIANS. 221 



portions of the egg, the cytoplasm for instance, for of course 

 there must be determinative factors (" determinants ") somewhere 

 in the ovum, probably in the nucleus which differ from one 

 another in kind. 



However in the phyla named the localization of morphogenic 

 substances in the cytoplasm is sufficiently definite to warrant a 

 comparison of one group with another. In all the ascidians 

 which I have studied and apparently in all which have been 

 studied hitherto, the type of localization is the same. Further- 

 more there is good reason for supposing that this type is essen- 

 tially like that of Amphioxus and Amphibia (on this subject see 

 Conklin, 1905). Judging from the work which has been done 

 on the organization of the egg in other phyla, this chordate 

 type is very distinct from that of annelids, mollusks, nemerteans, 

 echinoderms, nematodes or ctenophores. In fact it seems neces- 

 sary to recognize several distinct types of localization. 



If one has regard only to the localization of the substances of 

 the germinal layers there is considerable uniformity among most 

 metazoa in their pregastrular stages. In almost all cases the 

 ectodermal substances are localized in that hemisphere of the egg 

 which is nearest the polar bodies, and in this the ascidians are no 

 exception to the rule ; in many cases the mesodermal substances 

 are at first localized at the opposite pole, though only among the 

 echinoderms (Strongylocentrotus, Boveri, 1901) is this localization 

 persistent ; among annelids, mollusks, and ascidians the meso- 

 dermal substances early move from this pole to the posterior side 

 of the egg. 



However in the localization of specific organ bases there are 

 many notable differences among these phyla. In this regard the 

 annelids and mollusks and perhaps the nemerteans belong to one 

 type, the chordates, nematodes and ctenophores to entirely dif- 

 ferent types ; in fact the localization of organ bases in the ascidian 

 egg does not resemble that in the other phyla named any more 

 closely than does the localization of the larval or adult organs of 

 these phyla ; indeed, the principal chordate features are already 

 represented in the ascidian egg by characteristically localized 

 organ bases as early as the 2-cell stage. 



Since the time of Cuvier the principal criterion of homology 



