124 



through a pore than it finds in streaming in. For we found that 

 in carefully mounted preparations where all the choanocytes re- 

 mained fixed in their place, these cells surround the pores closely 

 and are placed, not exactly perpendicular on the wall, but some- 

 what oblique, so as to narrow the cloacal opening of the pore. 

 We found this to be the case in a flat, stretched piece of Leu- 

 cosolenia. Their position must be all the more oblique in the 

 living state if the wall is of course not flat but concave. If 

 therefore in the cloaca the pressure of the water becomes higher, 

 the collars of the choanocytes will become somewhat inflated and 

 the pore will be narrowed. If, on the contrary in the cloaca the 

 pression of the water in the neighbourhood of a pore is lessened, 

 water can easily flow in through the pores; the choanocytes with 

 their collars thus act as valves. Now by the irregular motion of 

 the flagella the pressure on the wall of the tube, which by the 

 spicules is kept rigid, is continually changing. If the pressure 

 becomes higher, this is of little eff'ect, but if the pressure becomes 

 less, water will flow in through the pores, as long as they are open. 

 The sponge will thus suck in water, which will leave the body 

 again through the osculum". 



Thus Yosmaer's and Pekelharing's theory. Then the investi- 

 gators call attention to the fact that the arrangement of the canal 

 system of the sponges becomes more and more appropriate accord- 

 ing as one examines higher developed forms. 



Personal Eesearcli. — How much this theory of Vosmaer and 

 Pekelharing's may correspond ivith their observations and hoiv well 

 it may make us understand that with such a movement of the water 

 a great number of foodparticles are brought within the reach of 

 the choanocytes, it could not quite satisfy me from the beginning. 

 It appeared very unlikely to me that the flagellar motion of 

 the choanocytes should principally be quite different from that of 

 the unicellular Flagellata eg. the Choanoflagellata, tvhich for the 

 rest remind us so much of the choanocytes. This supposition was 

 the more tempting, because it would make us quite independent 

 of the synchronism (whether existing or not) and of the direction 



