192 W. I. Robinson, 



3. The septa separate from their medial junction and in the 

 central area thus formed a columella develops as a vesicular 

 mass of calcium carbonate. The septa are not wholly discon- 

 nected from the columella, but their reduced edges form a fringe 

 of lamellae around it. The columnar vesicles are represented at 

 first by simple arched tabulae. This stage enters at the time when 

 about twenty septa have formed. 



4. The septa become completely separated from the columella. 



This is an example of an essential columella which shows 

 even in old age its origin in central tabulae and vesicles, without 

 any obscuring of them by a later secondary infilling of calcium 

 carbonate. Here too, however, there is a stage in which the 

 columella may be called parietal, although many septal ends 

 take part in its formation rather than one as in Lophophyllum, 

 or a few as is probably the case in Cyathaxonia. 



Both essential and parietal columellas are the results of an 

 invagination of the basal disk. It is not even necessary to 

 postulate an increase of secretion of calcium carbonate at the 

 center to account for the rapid upbuilding of this part of the 

 skeleton, since a central invagination in itself tremendously 

 increases the area of lime-secreting tissue. 



In the cases mentioned above the columella is evidently the 

 result of such an invagination of the basal disk. This may have 

 been in response to an upward pull at the center such as would 

 result from an attachment of the lower ends of the mesenteries. 

 In the case of the lamellar parietal columella of Lophophyllum 

 the force acted most eiTectively at two opposite points of the 

 polyp, causing an elongated invagination coinciding in direction 

 with a septal element. This led to a continuous lamellar plate, 

 the product of the fusion of a septum and the columella proper. 

 The septum involved is the counter septum, a fact suggestive 

 of the opinion held by various authors that the major septa lie 

 in the direction of the directive mesenteries. 



While the formation of all columellas is probably caused by 

 some physiological necessity, the result of which we see in an 

 invagination of the basal disk, there is yet a real distinction in 

 a morphological sense between parietal and essential forms. 

 The distinctive feature is the amount of influence the arrange- 

 ment and form of the septa exert upon the columella of the 



