106 THE ANATOMY OF INVERTEBRATED ANIMALS. 



ment take place, while the ova are still imbedded in the body 

 of the sponge. 



Metschnikoff ^ has recently described the development of 

 Sycoti cillatuni. The ovum, after impregnation, becomes a 

 morula, with a central cleavage cavity or blastocoele. But 

 the blastomeres of the two halves of the morula take on dif- 

 ferent characters — those of the one half elongating and 

 acquiring flagelliform cilia, while those of the opposite half 

 remain globular and develop no cilia. The latter now coa- 

 lesce into a syncytium, and develope spicula, while the layer 

 of ciliated cells becomes invaginated within the syncytium. 

 More usually, however, it appears that a gastrula is formed 

 by invagination of the morula, the ectoderm of which has the 

 structure of the endoderm of the adult, while the cells of the 

 endoderm, or lining membrane of the gastric cavity, are de- 

 void of cilia. The embryo quits the parent, propelled by the 

 flagelliform cilia which cover the outer surface of the ecto- 

 derm. After a time, it fixes itself by the closed end ; the 

 fiagella of the cells of the ectoderm are retracted, the cells 

 themselves become flattened and coalesce so completely that 

 their boundaries cease to be distinguishable, and the ectoderm 

 passes into the condition of a syncytium. At the same time, 

 the cells of the endoderm multiply, elongate, and take on the 

 form which characterizes them in the adult. In this state 

 the young sponge is termed an Asciila. The transition to 

 the final condition is effected by the development of the spic- 

 ula in the syncytium and the separation of some of the con- 

 stituent cells of the syncytium to form the inhalent pores. 



In the simplest Calcis]Jongice, forming the family to 

 which Haeckel applies the name oi Ascones^ the wall of the 

 ventriculus is thin, and the pores open directly into the ven- 

 tricular cavity ; but in another family, the Leuco7ies, the syn- 

 cytium becomes greatly thickened, and the pores are conse- 

 quently prolonged into canals (which may be ramified and 

 anastomose), connecting the ventriculus with the exterior. 

 The endodermic cells, which in these, as in the Asco7ies, at 

 first form a continuous layer, are eventually restricted to the 



1 " Zar Entwickelungs-cfoschichte der Kalkschwammc." {ZoiUdiriftfur 

 wissenschqftliche Zoologic^ Bd. xxiv. ) F. E. Schuke, so far as I follow Hae- 

 ckel's account of his recent onservations ('•'■ Die Gastrula und die Eifurcbung 

 der Thiere,'' p. 158), agrees with Metschnikoff as to t'ne first sta'^es of develop- 

 ment, but differs in regard to subsequent stances. Haeckel withdraws his ear- 

 lier account of the formation of the gastrula by delamination, or splitting of the 

 walls of an oval shut^?aww/a-3ac into two layers, and the subsequent opening 

 of the planula at one end. 



