118 M. K. A. Zittel on Fossil LitMstidce. 



group ; and where reproduction takes place by gemmation, 

 each branch possesses the same number of tubes as the 

 parent body. If, therefore, we are to regard each of the 

 above- described tubes as a separate stomachal cavity (as we 

 are justified in doing, seeing that they undoubtedly serve as 

 efferent canals) , the sponges belonging to this category present 

 examples of " syndesmotic " forms, in which each "person" 

 is capable of existing only in conjunction with several others. 

 The genera Jerea^ Thecosiphoniaj Polyjerea^ Marginospongia, 

 Stichophyma^ Jereica^ Turonia^ Doryderma, Carterella, &c. 

 serve as examples of this phenomenon. 



The question of individuality becomes still more difficult in 

 the cup-shaped and vase-shaped sponges. In these the wall 

 encloses a central space which is very wide above and nar- 

 rowed like a funnel below, the interpretation of which as a 

 stomachal cavity is certainly open to doubt, although nume- 

 rous similar radial canals of the same structure and direction 

 open into it. In many cases the oscula of these radial canals 

 attain a considerable size, and receive, on their own account, 

 the access of special lateral canals ; so that they themselves 

 play the part of flues or stomachal cavities, and the whole 

 sponge-body, like that of the common sponge, may be con- 

 veniently regarded as a composite stock. As, however, young 

 stocks possess the same cup-shaped or vase-like form as the 

 full-grown ones, as, further, the development of one of the 

 above-described oscula into a distinct stock resembling the 

 parent body has never been observed, and as these secondary 

 stomachal cavities at the same time also act as radial canals 

 of the whole colony, I leave the question of individuality 

 undecided, characterize such " strobiloid stocks " as simple 

 sponge-bodies, and place them in opposition to the " compo- 

 site " ones, in which, by gemmation in various ways, several 

 such strobiloid individuals of concordant habit are united into 

 a colony. Here, therefore, as among the Hexactinellidse, we 

 probably have polyzoic forms which in their external appear- 

 ance resemble a single individual, and, in a certain sense, are 

 equivalent thereto. This conception finds further support in 

 the fact that sometimes in one and the same genus the central 

 cavity decreases in dimensions and becomes gradually con- 

 verted into a wider or narrower funnel, the interpretation of 

 which as a stomach can hardly be doubtful. In the case of 

 the funnel-shaped and vase-shaped forms, moreover, one is 

 always in face of the dilemma, whether the central space is to 

 be regarded as the common efi'erent orifice, and the canal- 

 system is to be conceived as a unitary and coherent system, or 

 whether each large osculum, with the canal belonging to it, 



