439 



2. A Speculation on the Phylogeny of the Hexactinellid Sponges. 



By E. A. Minchin, Zoological Laboratory, University College, London. 



(With 2 figs.) 



eingeg. 1. Dezember 1904. 



By the classical memoirs of Schulze, supplemented by the more 

 recent investigations of Ijima, the body-structure of the Hexactinellids 

 has become thoroughly well known in spite of their deep-sea habitat. 

 The body-wall of a typical sponge belonging to this class consists of five 

 layers, namely, proceeding from without inwards, [1] a dermal membrane, 

 [2] a subdermal trabecular layer, [3J a layer of flagellated chambers, 

 [4] a subgastral trabecular layer, and [5] a gastral membrane. To these 

 facts, first made known by Schulze [10], Ijima [3] has added the dis- 

 covery, of most fundamental importance in the theoretical interpretation 

 of the sponge-body, that all these layers, with the exception of the 

 chamber-layer, consist of a uniform type of trabecular tissue representing 

 the dermal layer of the sponge, while the gastral layer builds up the 

 chambers. Thus as in all other sj^onges, the body is made up of the 

 dermal and gastral layers, together with a certain number of undiffer- 

 entiated cells or archaeocytes, from which are furnished the wandering 

 cells (amoebocytes) and the reproductive cells. 



It is not necessary to make further reference to the gastral layer, 

 the cells of which form, as in other sponges, an epithelium of collar-cells, 

 remarkable in many points of structure , but not calling for special no- 

 tice here. The dermal layer, on the other hand, presents features which 

 are unique in the whole group of sponges, and probably of a primitive 

 type in the phylum. In the first place, according to the Ijima's obser- 

 vations, the dermal layer shows no differentiation into epithelial and 

 skeletogenous layers, as in other groups of sponges, but the cells ana- 

 stomose with one another to form a continuous and uniform system of 

 trabeculae, between which the currents flow, and in which the spicules 

 arise. In short, the inhalant and exhalant canal-system is not formed 

 in Hexactinellids by definite canals lined by a flat epithelium, as in the 

 Calcarea and the Demospongiae, but consists simply of the meshes of 

 the trabecular system. Secondly, the cells of the dermal layer do not 

 secrete in Hexactinellids any gelatinous ground-substance or mesogloea 

 as in every other known sponge. Thirdly, the trabeculae of the dermal 

 layer are developed not only on the exterior of the gastral layer, but 

 also to an equal extent towards the inner side of the gastral layer. We 

 thus find a condition which, to anyone brought up on the ordinary or 

 text-book notions concerning the structure of the sponge body, may 

 seem at first sight very paradoxical; namely that the gastral layer, the 



