186 I. IJIMA : HEXACTINELLIDA. I. 



situated in the outermost part of the parenchyrae under the 

 dermal layer. On a priori grounds, it seems to me nearly certain 

 that that Anlage is, or is derived from, an aggregation or congeries 

 of archaeocytes, much the same as tliat which is so commonly 

 found in the deeper parts of probably all Hexactinellida after 

 maturity. The said Anlage is incorporated in the newly formed 

 bud, with perhaps a portion of the maternal dermal layer in 

 addition ; and, from it should originate b}^ differentiation most of 

 the different soft parts, or at any rate the flagellated chambers 

 of the progeny, — a fact, which is readily conceivable on account 

 of the blastomeric character I ascribe to the archa30cytes 

 (pp. 165, 172). Be that as it may, there stands nothing in the 

 way, so far as we know, of calling the above mode of reproduc- 

 tion in R. mirabilis asexual budding,* in which process the 

 archoeocytes seem to play a very important role. 



*Froin the character of the organization of tlie Hexaclinellid boily, it is not to he 

 wondered at, if a similar budding should take place in the interior of the sponge-wall, in- 

 stead of, or in addition to that, on the external surface. As an indication of such an internal 

 budding is perliaps to be regarded a remarkable case oC Sliurocahjplus glabcr Ij , in which I 

 found an innumerable number of small whitish bodies, distributed througliout the entire 

 parenchymal mass. On close examination they proved to consist of small, thick-rayed 

 liexactins fused together into an irregular framework, which was traversed by a few parenchymal 

 spicules. The framework, which I call the hasidictyonalia, is undoubtedly the same as that 

 described by Schulze ('99, P- 64; PI. XIV, figs. 2-6) from the bnds on the prostalia lateralia 

 of It. mirabUi-<. It may safely be concluded that a brood of the young had fixed themselves 

 on tlie parenchymal spicules of the aforesaid specimen of <S' glaher. The young may have 

 been asexually produced exactly as those of M. mirabilis, though not nece.ssarily so. In my 

 opinion, the hasidictyonalia is a structure which is formed quite generally in Rosscllids and 

 in certain other Lyssacine families evidently under the influence of the hard substratum 

 with which the sponge comes in contact at its base. It has hitherto been generally overlooked, 

 though the particularly small-meshed and most superficial layer of it has long been known 

 through Schulze. The little mass of hasidictyonalia in holh B. mirabilia and .S'. gluber la to 

 be regarded as having arisen in relation to tbe old spicules on which the yomig had attached 

 themselves. It may therefore be equally developed in all the young irrespective of these 

 having been produced asexually or from fertilized ova. However, under the assumption that 

 the brood in the above case of S. glabcr oiiginated from eggs, it seems somewhat strange that 

 the ciliated larvie were prevented from lieing st't at large during the free-swimming period.— 

 Of several R. yhibcr examined by me, the above was the only specimen in possession of a 



