IV] LARVA OF POKIFERA 89 



inhalant or afferent (Lat. ad, to; fero, I carry) canals which 

 intervene between them and into which the pores open (Fig. 39). 



When a sponge becomes still more complicated the central 

 cavity becomes broken up into a series of branching canals, which 

 are termed exhalantor efferent, and the flagellated chambers become 

 small and rounded (Fig. 38), each often connected only by a single 

 opening or prosopyle (Gr. Trpoo-w, forwards; -rrvXrj, a gate) with 

 the afferent system of canals. Numerous oscula are found in one 

 sponge mass, so that no pretence of discriminating the individual 

 can be made. 



A still further complication arises from the presence of sub- 

 dermal spaces. These are wide cavities immediately beneath the 

 surface of the sponge into which the pores open and from which 

 the afferent canals take their origin. In this way a rind or crust 

 of the sponge can be separated from a deeper part containing the 

 flagellated chambers. 



The larvae of sponges are best understood by a short description 

 of the simplest form, viz. the larva of Oscarella. This 

 has the form of a simple hollow sphere of ciliated cells 

 like the planula of Coelenterata in its first stage. The cells at one 

 pole lose their cilia, become pigmented and granular and then the 

 larva fixes itself by the ciliated pole. The whole animal flattens and 

 the granular cells extend over the ciliated cells, which become tacked 

 into the interior and there arranged as an inner lining to a cavity. 

 The flagellated chambers of the adult arise as small pocket-shaped 

 outgrowths from this cavity and the osculum is a later perforation. 

 The ciliated cells are eventually restricted to these chambers, where 

 they form the choanocytes and all the rest of the sponge is formed 

 from the granular cells. 



Other larvae differ from that of Oscarella in the early multipli- 

 cation of granular cells which form a solid mass at one end of the 

 larva, and often, indeed generally, this mass is of such extent as to 

 project into the interior. To compensate for greater dead-weight, 

 so to speak, the ciliated layer the locomotor organ of the larva 

 becomes extended so as to surround the granular material, so that 

 we are presented with the remarkable phenomenon of the internal 

 layer of the larva bursting forth and becoming the outer layer of 

 the adult. This is the case in the larva of Leucosolenia. In the 

 larvae of other calcareous sponges, the ciliated cells at first 

 surround the granular cells, but the latter are afterwards exposed 

 and the larva in this form has been called an amphiblastula. 



