53 



blue, for it staius tbe collars aud flagella. The preparation is 

 either washed iu water or af ter the blue tiuction treated with 

 ammonium-picrinate. The advantage of the latter is that the 

 colour can be better kept, However , it is at auy rate advisable 

 to study the preparations at once, for they always lose iu di- 

 stiuctness after some days. This is true, both for the isolated 

 cells and for sections. It is best to study the preparations in 

 water , diluted glycerine or acetas kalicus. Pure glycerine or 

 balsam are too refringent and render the collars less conspicuous. 



The highest power is necessary to get a fair idea of the re- 

 markable choanocytes. We used a microscope of Zeiss with the 

 new apochromatic leuses, viz. oilimmersion 1.40/3.0 with eye- 

 piece n° 12 or 18. 



It is only by this greatest care of preparing and by observing 

 the objects at this magnifying power of 1000 at least, that the 

 questioned details can be made out. The slightest omissiou or care- 

 lessness always resulted in a certain shrinkage or shrivelling up. 

 It seems to us certainly worth notice that iu such cases we saw 

 exactly the things as stated by the believers in Sollas's membrane. 



The sponges we could observe are : 



Halichondria panicea from Helder. 



Spongilla lacustris from the neighbourhood of Utrecht. 



Sycon dliatum from Bergen-op-Zoom. 



Y. Our preparations. 



If the sponges were thus treated with the utmost care and 

 studied with oil immersion , we could never see a tracé of Sol- 

 las's membrane. We often saw flagellated chambers with a net- 

 work iu the middle , exactly as this is described and illustrated 

 by Sollas, Dendy and Delage. In all those cases however it was 

 beyond doubt that the section was not so well preserved as 

 others. In all those cases the collars were quite irregular, folded 

 or shrivelled , and a good deal smaller than in our best prepa- 

 rations. In all those cases it were not only the choanocytes 

 which were smaller, irregular or in some way shrivelled, but the 



