153 



If one now inquires after tlie morphological meaning of this 

 layer of apparently undifforentiated flowing plasma, I must declare 

 that for this moment I don't dispose of sufficiënt data to answer 

 this question (but see Appendix). Considering the above mentioned 

 (p. 149 sub 1 6'), one is inclined to look upon it as consisting of 

 amoeboid cells (one or more for each flagellated chamber). But 

 one should not forget that that state mentioned of the sponge, 

 tested as ravel preparation, answers to what we got to know as 

 the state of the normal intact sponge on p. 149 sub 3. There it 

 proved that the canal walls in general, not especially the exterior 

 covering of the flagellated chambers, were loaded with carmine. 

 So it is quite possible that the amoeboid cells, treated on p. 149 

 sub 1 c, have simply been canal-wall-cells, and have not had anything 

 to do with that covering. The same counts for what follows. 



I killed a living sponge preparation, that, according to obser- 

 vation, had reached in a carmine suspension a stage as given on 

 p. 149 sub 3, in osmic acid and afterwards had the tissue mace- 

 rated in water, to finally study the separate cells under the 

 microscope. ïhe carmine now proved to be present: l^t in a great 

 number and in fine grains within the choanocytes 2°'i in a great 

 number and in big grains or conglomerates within very irregu- 

 larly branched amoeboid cells. These amoeboid cells can take the 

 most simple up to the most fantastic shapes ; one sees isolatcd 

 cells with long and broad protrusions, even extended to mem- 

 branes or stretched as bands (of course all pseudopodial processes), 

 exactly as was observed in living preparations as the apparently 

 undifforentiated plasmic substance; one sees cells with the ap- 

 pearance of a hollow tube, which evidently lined a canal, with 

 plasmic „bridges" and even membranes extended in the lumen. 

 All of them contain carmine. These amoeboid cells now carry 

 an often clearly visible nucleus, little or no symbiotic algae and 

 sometimes detritus ; they are rather numerous. But carmine is 

 hardly ever to be found in the ordinary amoebocytes with a great 

 number of symbiotic algae. 



One sees the striking conformity with what was found on 

 p. 149 sub 1, 2, 3. So the apparently undifforentiated plasma lining 



