239 



Two tliiugs are now generalij accepted , viz. : — 1". That the 

 simplest Calcaria homocoela represent the niost primitive canal 

 system , the ground type of which is a siraple sack ; and 2° that 

 the canal system of all Hexactinellids can be likewise derived 

 from a sack-type. On the other hand we see how , step by step 

 the complieated canal system of the Leuconidae can be derived 

 from the simple one of the Asconidae. 



Mr. Dendy's interesting paper has showu how manifold are 

 the ways in which the canal system of the simplest Leucosolenia 

 may become complieated, and may give rise to very different 

 things. It seems to me that one line of modifications tends to 

 the formation of a main central cavity, opening with a large 

 aperture, and in which smaller and narrower diverticles debouch. 

 We see this go on gradually. For in the type B. e. g. there is a 

 great irregularity ; only there is a central tube which is incont- 

 estably the main one , although other tubes may still be pretty 

 wide. In the most remarkable L. tripodifera however the difference 

 between the main tube and its diverticula is much more conspi- 

 cuous and there is much more regularity. 



I will for the moment not discuss how far the homologous 

 structures in the simplest Ascon and in the more complieated 

 forms like, say L. stipitata may extend. But I daresay that from 

 an anatomical point of view the main tube of L. stipitata is 

 aequivalent to the main tube of L. tripodifera. 



In the developement of Sycon we see that the following sta- 

 ges are passed. First the Olynthus-stage , wherein the central ca- 

 vity has no diverticula, then the stage wherein shallow diverti- 

 cula begin. And even in a not too old specimen we see fullgrown 

 chambers in the lower parts of the sponge, very shallow depres- 

 sions or ingrowths in the upper parts and transitions between both, 



What is now the chief difference between L. tripodifera and 

 a Sycon? It is this that in the former the whole inner surface 

 is lined by choanocytes, and that in the latter only part of it 

 (viz. the diverticula or flagellated chambers) is so lined. I do be- 

 lieve that the original flagellated cells of the inside of the Sycon- 



