170 I. IJIMA : HEXACTINELUDA. I. 



the groups, if already formed, are alwaj'S quite small and not 

 numerous. So that, it may safely be concluded that their increase 

 in number and their massive development take place along with 

 the sponge's advance in age. 



I believe the congeries of archoeocytes are of quite general 

 occurrence in the Hexactinellida. At any rate, I have met with 

 them, as they occur in small patches on the chamber-wall, alike 

 in a number of genera representing the different Lyssacine and 

 Dictyonine families, and it seems to me somewhat strange that 

 F. E. ScHULZE has omitted to indicate them in so many of his 

 figures of the chambers in the Challenger-Report (except in his 

 pi. LXII, figs. 7 & 8), though they were evidently seen by him 

 in some cases (see below). 



The larger arch?eocvte masses seem to have been noticed bv 

 some of the earlier investigators. Bowerbank ('62, p. 817) 

 mentioned that ' Iphiteon panicea Yal.' propagates by means of 

 'gemmules' which he figured badly (/. c, pi. XXXIV, figs. 17 & 

 18 ; also Mon. Brit. Spong., vol. I, pi. xxv, figs. 340 & 341). 

 Later, the same writer ('75, p. 506; pi. LVI, fig. 3) again found 

 in E. aspergillum a considerable number of ' gemmules ' dispersed 

 amidst the tissues and which were of various sizes and closely 

 resembled the same organ found by hira before in ' Iphiteon 

 panicea.^ It seems to me probable that the ' gemmules,' in E. 

 aspergillum at any rate, are the masses of archjeocytes, although 

 nothing respecting their composition can be gleaned fi'om the 

 works of Bowekbank:. 



Marshall ('75, pp. 153, 157) described peculiar oval bodies 

 which were found in Dactylocalyx, Sclerothamnus and Hyalonema 

 and were considered by him to be the same as Bowerbankian 

 gemmuhe. Here again I hold it likely that Marshall had before 



