16 THE VOYAGE OF H.M.S. CHALLENGER. 



Gonocytes (Eggs and Sperm). — It was very important to demonstrate that our 

 Deep-sea Keratosa develop eggs in the mesoderm, in order to show that they are true 

 sponges, and not large-sized Ehizopods. At first I vainly searched for them for a long 

 time ; but finally I succeeded in finding eggs in single specimens of all four families — in 

 Ammolynthus prototypus among the Ammoconidas (PI. VIII. fig. lC, e), in Psammina 

 plakina among the Psamminidse (PI. VII. fig. ID, e), in Psammophyllum flustraceum 

 among the Spongelidae (PI. V. fig. 5, e), and in Stannophyllum globigerinum among the 

 Stannomidse. The eggs were in all these cases of the same indefinite form and size as 

 in the other Keratosa, where they are described so well by F. E. Schulze and others. 

 They lie scattered in the maltha of the mesoderm, and exhibit always the large, clear, 

 subspherical nucleus, with a dark nucleolus, surrounded by the granular protoplasm. 

 The earliest stages of the eggs could not be distinguished from amoebocytes. 



It was not possible to distinguish spermatoblasts or ripe sperm in any of the Deep- 

 sea Keratosa, but considering the difficulties in showing their presence even in living and 

 well-preserved sponges, it is easy to conceive that they were not recognisable in our 

 insufficiently preserved spirit-specimens. 



Canal-System. 



The characteristic gastrocanal-system of the sponges exhibits, as is well known, 

 a great many modifications, which may be disposed in a few main forms or 

 types. In my Monograph of the Calcisponges (1872) I had distinguished three such 

 types, viz.: — 1. The Asconal-type (Ascon, Leucosolenia) ; 2. The Syconal-type (Sycon, 

 Sycandra); 3. The Leuconal-type (Leuconia, Leucandra). Vosmaer, in his recent 

 work on the Sponges (Bronn, 1887), has adopted these three types, and added a fourth 

 type, represented by Aplysina, the common sponge, Euspongia officinalis, &c. (loc. 

 cit., p. 144) ; this may be called shortly the Aplysinal-type. 



Two of these four principal types are represented among the Keratosa of the deep 

 sea. The canal-system of the new family Ammoconidse (PI. VIII.) is constructed on 

 the Asconal-type ; that of the three other families (Psamminidas, Spongelida3, Stanno- 

 midaj) follows the common Leuconal-type. The two peculiar types of canal-system 

 which we call the Syconal-type and the Aplysinal-type, do not occur among the 

 Keratosa here described. 



The difference between the simple Asconal-type of the Asconidse and the complex 

 Leuconal-type of the Leuconidse (and of the majority of all sponges) is so important 

 that many recent authors have adopted the separation of Polejaeff, who divides 

 the Calcarea into two orders, Homoccela (Asconiclae) and Heteroccela (all the other 

 Calcispongise). Employing the same principle in the Keratosa, we should divide 



