BY R. VON LENDENFELD, PH.D. 513 



course increased, so that it also appears advantageous that the 

 thickness and strength of the main fibres should be increased. 

 The spongiolin produced at one and the same time will be of the 

 same nature, so that the layers formed by it are not rendered 

 visible. For this reason the rapidly formed connecting fibres do 

 not show any stratification. 



The spongiolin produced at difierent times is slightly different 

 so that successive layers of spongiolin formed at greater intervals 

 of time will show their limits. This is the case in the main fibres. 



The axial thread is very well developed in the connecting fibre 

 and clearly visible, it consists of a granular mass and has a thickness 

 of 0*0004 mm. The connecting fibres join generally in such a way 

 that three fibres radiate out from one point. On close examination 

 it appears that two of these are portions of one and the same 

 continuous fibre with a continuous axial thread, whereas the other 

 grows out from it at the side. The axial thread of the latter 

 is not connected with the continuous axial thread of the former. 

 It terminates moreover with a trumpet shaped extension on the 

 surface of the other fibre. In the same way also the axial 

 threads terminate, where the connecting fibi-es join the main fibre. 

 This shows clearly that the three fibres radiating from one joining 

 point never are equivalent, but that the one is a primary fibre to 

 the side of which the other has attached itself afterwards and 

 therefore appears as a secundary. Light is thrown by this peculiarity 

 on the growth of the connective fibres. 



In the description of the skeleton I have mentioned that the 

 connecting fibres terminate at the sides of the grooves in the 

 skeleton with sharp points. It is always the primaries which 

 terminate in this remarkable manner (Plate 37, fig. 1, b.) The 

 points are very shai'p and abrupt so that the contour is somewhat 

 similar to a Roman short sword's point. The axial thread termi- 

 nates a little behind the point. 



The points stand very close together and the whole is a 

 most effective defensive arrangement. I am not aware that such 

 free and pointed terminations have been observed in other horny 

 sponge. 



