150 Natural History Bulletin. 



tion of calcareous salts in a sheath which is more or less 

 adherent to the horny axis. The difference between them is 

 in the form in which these salts are deposited. In P. dichot- 

 oma the deposition is in the form of the vermiform bodies and 

 in E. tourneforti it is in the form of spicules. 



The ending of the axis C3'linder in this species furnishes a 

 typical illustration of the theor}^ which I have advanced as to 

 the endodermal origin of the axis. 



In PI. VIII, fig. 9, w'e have a sketch made from a dissection 

 now in my possession, of the tip of the axis. It is a cor- 

 rect representation so far as the relation of parts around the 

 axis termination is concerned, the body of the polyp and 

 external outlines of the calicle being diagrammatic. 



The axis ends in a rounded boss, behind which is a raised 

 ring at which the mesodermal sheath of spicules ends. 

 The longitudinal lines indicate the flutings made on this soft 

 sheath by the impressions of the primary water-vascular 

 canals. 



The tip of the axis is seen to project upward until it is actu- 

 ally -within the caliciilar cavitv oj^ the tei'niinal polyp. 



The endodermal lining of the calicle embraces the terminal 

 boss of the axis and these endodermal cells seem almost 

 certainly to be the ones from which the horny axis is secreted. 



I have been unable to ascertain with certainty the signifi- 

 cance of the raised ring below the base. It seems probable 

 that it was once the boss and is. destined to be flattened out 

 until it constitutes one of the lamellcC of which the axis is com- 

 posed. 



These lamellae are similar to those found in Miiricea for 

 instance, and illustrated in PI. Ill, fig. 5, a. 



If my supposition is correct we can easily account for this 

 laminated appearance by regarding each lamina as having 

 been a boss of soft, almost gelatinous tissue which has been 

 flattened upon the preceding one by the pressure arising from 

 the formation of a new boss above it. This will also account 

 for the fact that the laminae are thicker near the tip of the axis 

 than further down. 



