Hyatt.] 104 [March 6, 



the limits are in this direction. We therefore provisionally 

 assume, in accordance with Marshall's research, that the endoderm 

 extends into the incurrent canals beyond the ampullae, because 

 this view seems to be in accord with the general homologies of 

 the coelomic system in other branches of the animal kingdom, 

 and enables one to homologize the tubes in part at least with the 

 tubes of Coelenterata and Echinodermata. 



The possibility that the primitive cloaca may have owed its 

 origin to the pressure of water accumulated in the interior of the 

 ampullinula has been already advanced by Barrois, and this opin- 

 ion is sustained by the vase growths above described. The cloacal 

 tubes in adult sponges appear to be kept open by the action of the 

 more or less constant outflow from the interior, and extra growth 

 of the peripheral parts building up the vase or the cloaca is indi- 

 rectly due to the same cause. It is difficult to imagine the primi- 

 tive cloacal tube to be due to any external influences. If a gas- 

 trula is present, and the normal transformation of the blastoporic 

 area and its collar or velum (Lankester) into an organ of attach- 

 ment be also granted, then the breaking through of the orifice and 

 the formation of the cloaca at the free end would be an inevitable 

 result of any accumulation of water or fluid within. The orifice 

 would be formed as a simple opening. The next step would be 

 the invagination of the ectoderm which would arise as a secon- 

 dary ingrowth. The aspect and structure of the layers in the 

 peripheral cavities of Spongia leave very little doubt that the 

 true ectoderm extends to a certain extent internally. The color 

 and aspect of the ectoderm and derm in the superficial canals 

 is often precisely the same as upon the exterior, and it is evident 

 that these are formed by invagination or ingrowths of the ecto- 

 derm, though whether this layer extends inwards any farther or 

 not, may be considered doubtful. 



This view of the lining membrane of the canal does not 

 agree with that of any other investigator, but the histological 

 similarity of the lining membranes in the cloacal and sup- 

 ply systems or the series of transitional structures above given 

 cannot be accounted for unless the true ectodermal cells can be 

 considered as extending far into the interior in the cloaca, and 

 also in the supply canals, at least in full grown specimens. 



If our conclusion is trustworthy the invagination of the ecto- 



