49 



out by serial sections whether a canal is excurrent or incurrent, au exceedingly puzzling business, 

 which generally demands much time and much patience. 



In the shape of the mastichorions we have an index to detennine at which side the 

 water comes in and where it flows out. Pekelharing and I showed (1. c.) in which way the 

 choanocytes of the mastichorions may act as valves. According- to our views the opening which 

 is called apopyle must forcibly be the excurrent opening of a mastichorion. Consequently the 

 canal, which communicates with the apopyle, is the excurrent canal, that on the other side the 

 incurrent one. In a specimen with wide central canal (1365) I now tried to make out the course 

 of both the excurrent and incurrent canals and, as statecl above, I came to the conclusion that 

 the wide central canals communicate with the incurrent canals and that the excurrent canals 

 finally open at the sponge surface by means of a large quantity of small openings (PI. Vlij. 

 This involves that the wide central canal is incurrent, not excurrent. 



I need not say that it is only after long hesitation that I dare state this. Also I state 

 it with all possible reserve, being convinced that it is highly necessary that others confirm it 

 for this or lor other sponges. I can only say that in the case under consideration I cannot 

 explain things otherwise, but I admit of course that I can be misled by my preparations. I 

 don't wish to pretend yet, that in some sponges the wide central canal is surely incurrent, but 

 I do think it worth while to stimulate other spongiologists to pay attention to it and to settle 

 the question either in one sense or the other. I may add that I have myself tested other sponges 

 and that in one or two other cases I found similar things as in 5". purpurea. 



If I ara right that the course of the water really is as described for 1365, in which 

 case there is not a cloaca with wide mouth, does this throw over the hypothesis emissed by 

 Pekelharing and myself? By no means. For I believe that in Spirastrella purpurea the 

 water current is not or not always effected by the action of the flagella of the choanocytes 

 only. My conception of the large central canal is that it represents a reservoir, belonging 

 to the incurrent system. We saw in 1365 that the wall of the canal is provided with well 

 developed muscle cells, which for the greater part are arranged concentrically. If these cells 

 contract, the lumen of the canal will diminish. We likewise saw that the lumen at the periphery 

 generally is smaller, either because of narrower canals, or because it is contracted. Finally 

 we saw that the opening or openings may be shut. Suppose this to be the case ; if then the 

 muscle cells contract and therewith the lumen is lessened the water will be driven into the 

 smaller incurrent canals respectively the mastichorions etc. In this connection it is worth while 

 to remark that the total amount of mastichorions compared to the total sponge volume is 

 small (PI. VII, fig. 1). 



We arrivecl at the conclusion, that the large central canals and some of the peripheral 

 ones belong to the incurrent system and that, consequently, their narrow or wide openings at 

 the top of the processes are not "oscula". This does not involve that all the larger openings 

 we see in certain specimens are incurrent apertures. In specimens belonging to tropus glaebosa 

 we remark on the sponge surface a greater or smaller amount of comparatively large openings 

 (PI. I, fig. 4; PI. II, fig. 5; PI. V, figs. 6 and 7). Longitudinal sections through such openings 

 teach us that in these massive, more or less spherical specimens the greater part if not all of 



SIBOGA-EXPEDITIE VI « '. 7 



