BY R. VON LENDENFELD, PH.D. 297 



Osculum, whicli are cylindrical, but do not possess a collar, and 

 'which in every respect present transition forms between the 

 chamber epithelium, and the canal epithelium. 



Skeleton. 



The skeleton of the Halme Nidus Vesparum specimens observed 

 by me is always the same however different the sponge may appear 

 in shape and size. 



Main fibres and connecting fibres are, as stated above, in the 

 description of the genus, very different from each other. 



The main fibres grow up from a basal horny plate containing 

 much sand and tend upwai'ds remaining in or near the centre of 

 the sponge lamellae. They finally curve slightly outward and ramify 

 in a penicillate manner. The branches are as thick as the stems 

 and in their terminal portions more or less parallel. 



Comparing our sponge with others we notice that the ramification 

 of these main fibres is but slight and that the main fibres never 

 hardly seem to coalesce to form a reticulation. 



The main fibi'es are on an average 0'3 mm., thick. The greater 

 portion of them, the whole inner part (Plate XXVIII,, fig. 8, Plate 

 XXVII., fig. 7) is taken up by a dense mass of large sand granules 

 held together by a Spongiolin-cement. Outside we find a thin 

 horny layer on the surface of the sand. This is stratified and the 

 layers are visible because the successive strata refract the light in 

 a different degree. 



This outer coating or sheath is very thin, measuring only 0-01 

 mm., in thickness. Consequently the surface or outline of the fibre 

 appears very uneven as the projecting corners of the imbedded 

 sand grains are divided by indentures, which the outer horny layer 

 is not thick enough to fill up. 



These main fibres are connected with one another by connecting 

 fibres (Plate XXVII., fig. 8). These are very simple and as shown 

 in the figure mostly straight, vertical to the main fibres and only 

 very slightly ramified. 



