98 



Dr. J. E. Duerden on the 



upwards of tliirtj enlarged examples found; in only one or 

 two doubtful instances was but one mouth remaining. Evi- 

 dently therefore the increase of tentacles and mesenteries 

 early results in oral fission, if the two, indeed, do not proceed 

 pari jjassu. The plane of fission coincides with the median 

 or axial-abaxial plane, and in some rare instances the two 



IV 



Fig. 1. 



Transverse sertion tlu'ough the stoiuod?eal region of a projecting radial 

 polyp of JMadrepora. In relation to the axis of the branch the lower 

 aspect of the section is axial and the upper is abaxial. The two 

 mesenteries, IV, IV, forming a bilateral pair, are the dorsal direc- 

 tives, and mesenteries ///, III are the ventral directives. Mesen- 

 teries II, V and /, VI on each side constitute two anisocnemic 

 unilateral pairs. The polypal cavity is in communication with the 

 superficial canals by means of several radiating canals. The outer 

 polypal wall (coenosarc) is supported upon tw enty-four costal ridges. 

 In this and all the figures the three coslenterate layers are repre- 

 sented as follows : — The ectoderm by the conventional columnar 

 ejiithelium ; the mrsor/laa by a tliick black line ; and the endodcrm 

 as a clear layer. Owing to its narrowness the calicoblastic ectoderm 

 is unrepresented, the skeletotrophic mesogloea appearing to cover the 

 coralium directly. The skeleton or corallum is indicated by the 

 dotted areas. 



])olypal halves appeared to be in process of separation by the 

 ingrowth of the middle tentacles and septa. Increase in the 

 number of polyps in any colony of Madrepora by means of 

 fission must, however, be very insignificant compared with 

 the usual increase by coenosarcal gemmation. 



The early stage at which oral fission is accomplished in 



