406 Mr. W. J. Sollas on the 



them with the incurrent system. On the other hand they are 

 gradually prolonged into the small ultimate canals of the excur- 

 rent system (PI. XVII. fig. 25) ; they are the expanded ends 

 of these canals, which unite together into larger trabecular tubes, 

 having no direct communication with ciliated chambers, except 

 that furnished by these tributary ultimate canals. 



Herein lies the great distinction between the incurrent and 

 excurrent system. The tubes of the former communicate 

 directly at every part of their course with ciliated chambers ; 

 the tubes of the latter only communicate with the chambers at 

 the end of their ultimate ramifications, just as a tree only 

 bears leaves at the end of its twigs. 



This observation, in connection with the difference in the 

 mode of connexion (first pointed out by F. E. Schulze) of the 

 excurrent and incurrent canaliculi with the ciliated chambers, 

 is very suggestive. The cells of the ciliated chambers, toge- 

 ther with the epithelial lining of the excurrent canals, are the 

 adult representatives of the endoderm of the larval sponge ; 

 the epithelium of the incurrent tubes, together with the epi- 

 dermis, are the descendants of the original ectoderm. In 

 course of growth the ectoderm and endoderm have increased 

 more rapidly than the intermediate tissue, which F. E. Schulze 

 terms mesoderm ; and the result has been an involution in two 

 opposite directions — the endoderm developing like a racemose 

 gland in one direction, the ectoderm undergoing a simpler 

 involution in the other; such, at all events, appears to me 

 the origin of the canal-system in Isops and Oeodia. 



Our observations might, however, be brought into accord- 

 ance with Hackel's theory of the canal-system, if we consented 

 to regard our incurrent canals as forming an intervascular 

 system, and the excurrent only as a genuine gastrovascular 

 system. At the same time this is a purely theoretical view ; 

 and I cannot see how one reasonable man can blame another 

 for choosing to consider the canal-system of such a sponge as 

 Isops or Geodia as having a so-called "bipolar" arrangement, 

 which, as a matter of observation, independent of all theory, 

 it has. In saying this I am far from expressing any difference 

 of opinion from Hackel, whose general conclusions are clearly 

 in the main correct, but simply desirous of adding my testi- 

 mony to the value of Carter's observations, which are always 

 faithful and accurate, and worthy a more generous estimate 

 than that awarded them by his opponent. 



While speaking of the canal system I would take the 

 opportunity to point out the fact that the vesicular character 

 of the incurrent canals is of a totally different nature from that 

 described by Hackel as distinguishing his " blasenformige " 



