5i 



external surface at least at the top, and we learn by microscopical sections that the structure 

 is indeed the same. In other words, these cavities turn out not to be homologous with the 

 central canals of 1365, 1975 etc. but with the interstitial cavities of 931b etc.; they are, conse- 

 quently, cloacal cavities. In lact 944a has rather delicate processes 1 ), corresponding to those 

 of trop. digitata and tuèulifera, which ma) - fuse in a way that the space between them remains 

 larger than their own diameter. 



In many cases I found such interstitial spaces inhabited by an Ophiurid (PI. V, 

 fig. 8; PI. I, hg. 2; PI. IV, lig. 5). I am inclinecl to believe that this Ophiurid (Op/iiacfis 

 savignyi Muil. & Troschel) -) will be found, it not in every interstitial cavity, certainly in the 

 majority of them. Since we know how Eapagurits by its continual movements keeps open 

 a cavity in Suberites domuncula, it may be suggested, that in the same way Ophiactis is for 

 a good deal responsable for the large size of the interstitial cavities in Spirastrella piirpurca 

 (Cf. infra). 



To resumé: there are in Spirastrella purpurea large canals in the centre of the processes, 

 which are to be considered as incurrent water-reservoirs and which possess at the top of the 

 process an opening. This opening as well as the lumen of the central canal can be narrowed 

 or closed by contraction of muscle cells. Apart from these incurrent openings, other ones are 

 found in several cases : they generally appear as the mouths of interstitial cavities (grooves or 

 pits), sometimes as terminal openings of more or less cylindrical elevations, thus simulating the 

 incurrent apparatus, in fact belonging to the excurrent system. 



On the external surface of some specimens one sees, with the naked eye, irregular spots 

 (PI. V, fig. 23), which are more or less transparent. Microscopical sections show that we have 

 to do here with cavities, covered by a thin layer of tissue ; hence the semitransparent appearance. 

 Seen with low power in strong light numerous small apertures become visible (PI. V; figs. 21 

 and 22). These small apertures are either stomata or procts ; I hardly find a difference in size, 

 so that these are not distinguishable from each other. Especially not because the size varies 

 according to the state of contraction and in many specimens the apertures are invisible because 

 they are closed. I cannot teil whether in- and excurrent apertures are localised ; but it seems 

 certain that in case the sponge possesses an incurrent reservoir, nevertheless stomata may be 

 present at the outer sponge surface. At any rate the stomata lead by short narrow canals into 

 wider canals, more or less parallel to the surface; the latter probably correspond to "subdermal- 

 cavities". They communicate with similar superficial cavities and the deeper canals, giving off 

 branches to the mastichorions (PI. VI, fig. 1). These are situated in groups and are not very 

 numerous, often even rather scarce, except in incrusting (young?) specimens. The system is 

 euripylous (PI. VI, fig. 2). On the whole the excurrent part of the canal system is very much 

 like the incurrent one, as we have seen in the examples described above, viz. 1945 and 1365- 

 We find it so, mutatis mutandis, in all the different tropi. The stomata as well as the procts 

 communicating through short narrow canals with wider ones, which often unite with other similar 



1) The specimen shows these processes on the side opposite to the one which is photographed. 



2) Deteimined by Prof. Koehler. 



