135 



taken from Schulze (52) and already mentioned by Vosmaer 

 and Pekelharing (62), regards sponges which show the diplodal 

 type of canal-system, where in general there does not seem to 

 be such a great difference in width of the prosodi and aphodi, 

 as both are narrow. It now appears that here (in Chondrosia eg.) 

 the flagellated chambers are pear-shaped and that they only carry 

 choanocytes at the side, opposite to the „stem", while the „stem"- 

 side is covered by pinacocytes. Now the „stem" is formed by the 

 excurrent canal, while the incurrent canal ends among the choano- 

 cytes. It is a matter of fact that, with such a structure, the sepa- 

 ration of the flagellated chamber into the three pressure-zones 

 wanted is guaranteed. 



The importance (to the water-current within a sponge) of well 

 placed choanocytes, with respect to the apopyle of a chamber, 

 appears, however, even more clearly from what I have been able to 

 state on my living microscopic preparations of Spongilla. It proved 

 several times, when I had the chance to observe a flagellated 

 chamber with its excurrent canal exactly from the side, that here 

 too the choanocytes are almost exclusively attached to the wall 

 opposite to the apopyle (see also Fig. 55, after Vosmaer and 

 Pekelharing) and that, when moving, all flagella are directed 

 towards that opening, that even, sometimes, they are beating 

 with the tops outside the apopyle, so in the excur- 

 rent canal (Fig. 64, 73). The latter is of much importance, as 

 will soon appear. 



First something else. On p. 131 I mentioned that the collars 

 only left their flagellum uncovered for rather a small part at the 

 top (Fig. 59). The consequence will be, that the flagella only in- 

 fluence the water with a small part (the top outside the collar) ; 

 for it goes without saying that the water-circulation within the 

 narrow collar can not be important and, besides, will be closed 

 for the greater part in itself. Consequently, the choanocyte pos- 

 sesses in the length of the collar a very good means of regulating 

 the effect of its flagellar motion: the shorter the collar, the stronger 

 the effect. One might make the objection that energy w-ould then 

 be wasted by the flagellar motion within the collars. But con- 



