﻿DUERDEN] MOKl'IIOLOC.V OF CORAL POLYPS IO D 



not fortuitous in character, for in the same species free swimming 

 larvae have been found within the polypal cavity. Hence there is 

 little doubt that all corals are viviparous. It is somewhat remarkable 

 that though ripe polyps, and others charged with larvae, have been 

 abundantly met. none of the intermediate developmental stages be- 



Fig. 12.— Throe stages in the development of the twelve pairs of third-cycle 

 mesenteries. All the six pairs of primary mesenteries are now complete, and 

 the second-cycle pairs are all equal, but free from the stomodseum. In a a pair 

 of third-cycle mesenteries (in) has appeared on each side, within the exocoele 

 next the dorsal directives ; in b a corresponding pair occurs within the dorsal 

 of the two exocoeles of all the six systems, the order being from the dorsal 

 to the ventral aspect; in c another series of six pairs is commencing, situated 

 within the ventral of the two exocceles in each system. Here, again, the 

 growth in the dorsal region is in advance of that in the ventral. 



tween the egg and larva has been encountered within the polyp. 

 The only account of the segmentation and formation of the germ 

 layers available is that given by H. V. Wilson 1 for Mankind areolata. 



1 Journal of Morphology, vol. n, 1881 



