Madt-eporarian Genus Turbinaria. 509 



apparent periodical growth is specially marked in erect 

 fronds, because at the boundary line between the new and 

 the old a projecting ridge, studded along its edge with young 

 calicles, often forms a sharp contrast between the two. 



This appearance, however, is, in this case at least, entirely 

 delusive, and is due to the occasional streaming back of coral- 

 substance from the living on to the dead portion of the 

 corallum. Fig. 4 (PL XIX.) shows a portion of an erect frond 

 which has fortunately fractured tlirough one of these apparent 

 border lines between new and old corallum . The growth is seen 

 to be perfectly continuous, the lower part progressively dying, 

 on the left face (height c) faster than on the right {d). Two 

 floods of coral-substance have streamed down (to a and &), 

 but in neither case do they overflow the dead corallum, but 

 they submerge the living. Fig. 5 shows a calicle being over- 

 whelmed by finely reticulated coenenchyraa. A study of a 

 fracture passing through the edge of such an advancing flood 

 of coenenchyma shows that the calicles thus overrun are able to 

 work their way again to the surface. The earlier flood marked 

 l was so abundant that it formed the shelf shown in the figure, 

 along the edge of which a number of minute calicles appear. 

 The origin of these calicles I have not made out. I suspect 

 they are due to the secondary budding of the submerged 

 calicles which failed to break through the layer of coenen- 

 chyma which overwhelmed them. 



The Turbinarians, then, are no exception to the rule of 

 progressive dying down. The gradual character of this is, 

 however, obscured by occasional downflowings of coenenchyma 

 forming projecting ridges, which appear to indicate distinct 

 periods of growth. 



The continued downstreamings of the coenenchyma, de- 

 structive as they are to the lower polyps, clearly add to the 

 thickness of the basal portions of the corallum as the growing 

 edges of the fronds rise higher and higher. 



On the other hand, there are cases which can, I think, only 

 be explained on some theory of periodicity of growth. There 

 are specimens in the National Collection in which small points 

 of fresh growth are to be found on the edges of otherwise 

 dead coralla. Certain growth-forms, presently to be described, 

 seem to require such regular periods ; but in these cases the 

 new growth, without passing exactly through the early cup- 

 stage, repeats more or less independently the growth of the 

 old stock. 



There is one very remarkable specimen in the National 

 Collection in which a new cup develops from the margin of 

 an old one. The old cup has, however, been turned com- 



