G ORG ON AC E A, AND SOME NOTES ON THE SUBEROGORGIID.E. 227 



covered with an epithelium, which can not be anything else than the 

 Axenepithel. The development of this epithelium differs very much in 

 different colonies. In most cases it is limited nearly to the apex of 

 the axis, so that the cortical portion of the axis reaches close up to 

 the tip. Sometimes however I have met with such cases, in which 

 the epithelium reached nearly to the lower end of the uppermost 

 polyps. In typical cases, one of which is shown in fig. 1, the cells 

 composing the said epithelium at the tip of the axis are very tall 

 (0.02 mm.), but become gradually less tall below. The final fate of 

 this epithelium has not become quite clear to me. In one case met 

 with, the cells of the epithelium seemed to lose themselves in the 

 horny matrix binding the spicules of the cortical layer. The latter 

 develope even in the mesoglcea of the partition between two terminal 

 polyps. A little downwards the horny matrix appears between 

 the spicules, binding these to the central cord, while the epithelium 

 disappears, the cells becoming lost in the horny matrix (fig. 2). In 

 most cases this stage is not to be found ; probably it transpires 

 with rapidity. 



In the majority of the specimens examined, the Axenepithel is 

 very difficult to detect, owing to its limited extent at the tip of axis 

 and its proximity to densely crowded solenia of various caliber, but 

 especially to the shrunken state of the tissues, whereby the distinction 

 between the epithelia of solenia and the Axenepithel is rendered ex- 

 tremely difficult. 



Now a word about the mode of branching of the colonies. Gen- 

 erally the branching takes place near the tip of branches. However 

 it often happens that a branch seems to have been produced as a 

 lateral bud from a larger stem or branch. In examining the origin 

 of such a branch, I have found that the central cord of the branch 

 may be separated from that of the stem by a thin cortical layer con- 

 taining spicule-lacunae. It docs not necessarily follow from this fact ( 

 that the central cord is not an epithelial secretion, since the possibili- 



