394 H. N. MOSELEY. 



topora it becomes, when spirit specimens are decalcified, 

 beautifully injected with the resulting gas, and all its network 

 is then rendered plainly visible. Exactly the same result occurs 

 when specimens of Corallkim nihrwn are decalcified, the main 

 canal network becoming similarly distended and rendered con- 

 spicuous. 



On finding the above described structures in Seriatopora, I have 

 again examined some microscopic preparations of Pocillopora. The 

 polyps in Pocillopora appear, like those in Seriatopora, to have 

 only a single pair of very long mesenterial filaments, and these 

 filaments appear to correspond in position to the two long ones 

 in Seriatopora, that is to say, to belong to the central mesenteries 

 of the lateral chambers. I suspected long ago that Pocillopora 

 had only a single pair of mesenterial filaments in each polyp, but 

 I did not feel certain, because these filaments are not enclosed in 

 prolongations of the chamber-walls, as in Seriatopora, and they 

 are so excessively long that in decalcified preparations those 

 belonging to adjacent polyps become interlocked and supposed, 

 so as to render exact recognition difficult, and I did not devote 

 time to the point. As stated, there are no such conical prolonga- 

 tions of the walls of the inter-mesenterial chambers in Pocillopora 

 as exist in Seriatopora, but in the bottom of some of the calicles in 

 some Pocilloporas a pair of deepish pits may be made out in 

 corresponding position to the deep pits of Seriatopora, and evi- 

 dently serving for the reception of the single pair of long fila- 

 ments. An arrangement of rudimentary septa, closely similar to 

 that in Seriatopora, may also be made out with difficulty in some 

 Pocillopora calicles. I have not been able to find generative 

 elements in any specimens of Pocillopora which I have examined. 



There is a canal network in Pocillopora like that existing 

 in Seriatopora, but in the species which I have as yet examined 

 it difl^ers from it in that its vessels are wider and coarser, and 

 show a tendency to degenerate into lacunar spaces. 



The presence of the deep pits in Seriatopora for the reception 

 of the single pair of generative mesenteries and their hypertro- 

 phied mesenteries may possibly explain the pits occurring amongst 

 the septa of some palseozoic corals which may have had a similar 

 function. 



The indications so plainly marked in the coralla of both 

 Seriatopora and Pocillopora had already shown that in these 

 compound forms of the Madreporia at least there exists, as in the 

 Alcyonaria, a very definite and uniform orientation of the polyps 

 in each colony, in accordance with the bilateral symmetry and 

 dorsal and ventral diff'erentiation of their structure. This fact, 

 indicated by the arrangement of their septa, is fully borne out 

 by the structure of their soft tissues. 



