DEVELOPMENT OF HALISARCA LOBULARIS. 617 



produce a segmentation cavity and a blastula form. This may 

 not have been found necessary in all cases, and the solid 

 morula, like that of our Hal is arc a embryos, may represent 

 such an exceptional dispensation with the early blastula form. 

 From this point the course of development probably diverged 

 in different groups of Sponges. In some a hollow blastula of 

 flagellate individuals was produced, and this either owing to 

 osmotic action or unequal growth at diflferent parts of its peri- 

 phery became infolded at one point or more. The infolding once 

 commenced would proceed, pari passu, with the growth and 

 subdivision of the invaginated cells. If dependence is to be 

 placed on our Halisarca embryos one may suppose that several 

 foldings were in some cases produced. The sinuses of the 

 folds would probably serve as quiet recesses in which food 

 particles would be collected. Anyone watching an infusorian 

 at work must have been surprised at the few particles of food 

 it secures compared to the currents it sets in motion. There 

 is evidently here a great waste of energy, and possibly some 

 of this is spared by the formation of a gastrula. But the 

 currents produced by flagella in a sac open only at one end 

 would become less effective as the sac became more complete, 

 and would cease to be of any use in obtaining food as soon as 

 the gastrula attached itself to some foreign body by its oral 

 end. Hence the development of pores which restore commu- 

 nication with the exterior, though it is difficult to understand 

 how the pores arose.^ Subsequently the mesoblast would arise 

 by migration of the endodermic cells. 



In yet other groups of Sponges, probably the most numerous, 

 the blastula buds off internally cells which give rise to mesen- 

 chyme, comparable to the cytoblast of the ancestral infusorian 

 colony', and thus the ancestral form of the Sponges appears at 

 an early stage. The gastrula cavity is now formed by an 

 internal splitting of the mesenchyme, and this need not neces- 

 1 Meriiowski describes pores as perforating the wall of the blastula of the 

 embryo of Obelia ('Bull. d. 1. Soc. Zoo. d. France,' 1883, p. 27, Extrait). 

 As they, however, do not occur iu the gastrula, they cauiiot be homologous 

 with the pores of Sponges. 



