182 THE EVOLUTION OF THE META20A 



sarcosepta which protrude into the gastrovascular space. These 

 free margins have become thickened (in a transverse section 

 they appear completely round), and in the epithelium we find 

 not only gland cells but also cnidae. The curled margins can 

 grow locally into long filaments which can be pressed into the 

 free water through special openings that develop in the body 

 wall (cynclidae) to be used there as weapons (acontia). 



The transition to cormogony (the formation of colonies), 

 connected w ith the decrease of the size of the polyp individual 

 has led to a reduction of sarcosepta. This reduction was both 

 quantitative (the reduction and simultaneously the fixation of 

 the number of sarcosepta to six or eight), as well as qualitative 

 (small acontia, few^er filaments, etc.). In the skeleton forming 

 Madreporaria, the digestive surface has been secondarily in- 

 creased by means of sclerosepta which grow from below into 

 the gastral space. This reduction is coordinated with the re- 

 duction of the gullet pharynx tube. Sarcosepta are not present 

 at all in the solenia of corm -forming species and here we there- 

 fore have the opportunity to observe all possible transitions 

 in one and the same object, from a typically anthozoan in- 

 testinal tube, to the kind that can be found in Scyphozoa, and 

 finally to a perfectly smooth intestinal tube, a state that has 

 been reached in Hydrozoa. 



The reduction of the septal apparatus has gone even further 

 in scyphopolyps, parallel to the decrease of their size and to 

 their sessile way of life. In this way this reduction can now be 

 understood. The number of strongly diminished and cytologi- 

 cally simplified sarcosepta (they are also called taeniolae) has 

 been reduced to the possible minimum, i.e. to four, in an 

 increasingly developed radial symmetry. If we take the old 

 view of a progressive evolution along the line Hydrozoa -> 

 Scyphozoa -> Anthozoa, we find it impossible to explain how 

 the sessile polyp of Cnidaria could suddenly begin to form 

 sarcosepta. This situation becomes even more difficult to 

 understand if we deny to the scyphopolyp the character of 

 a special generation, as was done by Hyman. 



