510 Mr. H. M. Bernard on (he 



pletely over, with the result that its margin has curved 

 upwards all round, and at one point shot up to form a new 

 cup. ]n this case there was evidently no periodical dying 

 down of the first cup ; its normal course was interrupted by 

 an accident. 



I have distinguished eight principal growth-forms among 

 the Turbiuarians in the National Collection. In addition to 

 these eight there are a few specimens whose method of growth 

 comes under no heading ; and, whether they be normal or 

 accidental, we are not in a position to decide until the collec- 

 tion of the Turbinarians is more complete. Each of these 

 eight forms requires description. 



First TyjJe of Groicth : Crateriform. — This, as the name 

 implies, is a simple persistence of the early cup-form, not at 

 all or but slightly modified. As above stated, the existence 

 of large cups, in face of the great improbability of the budding 

 being sufficiently regular, compels us to attribute a classifica- 

 tory value to this method of growth. We cannot assume 

 that it is the result of mere favourable chance until by experi- 

 ment we have proved it. It seems further only natural that 

 of all the various normal growths of the Turbinarians some 

 species should depart less from the initial form of the corallum 

 than others. Be this as it may, until our knowledge of the 

 genus is much more extensive we have no other course open 

 to us than to assume the persistence of the cup-shape through- 

 out life to be a normal character distinguishing certain species 

 from the remaining members of the genus. 



The group is also practically of great use, inasmuch as all 

 cup-shaped coralla whose subsequent method of growth is 

 unknown may be provisionally placed in it. 



On the outside of the large cup (referred to on p. 508) at 

 various heights there occur several attempts to form small cups, 

 which are generally much distorted owing to theanglesat wliich 

 they project from the parent-stock. These are, I think, to 

 be associated with the streaming of the coral-substance. They 

 appear, at least in many cases, to arise where the downward 

 flow has been hindered by some obstacle. 



Second Type of Growth: Peltate. — As above stated, the 

 peltate growth may continue along two distinct linesof develop- 

 ment, according as the early disk has the edges tending to 

 fold upwards or downwards. We have here then, from the 

 nature of the case, possibilities for great variation in growth 

 in one and the same species, the peltate young form standing 

 on the border-line between the cup and the glomerate or 

 ratiier the flat encrusting type of growth. It is significant that 



