DEVONIAN ROCKS OF WESTERN CANADA. 49 



In such, cases, the moutli or calice of tlie corallites appears to be closed 

 "with a kind of disc, which is sometimes level with the general surface, 

 often depressed slightly below it, and sometimes elevated in the form 

 of a rounded boss. 



Tn a very large number of specimens, the epitheca has been more 

 or less denuded over parts where it originally existed. In such cases 

 it is mostly only the epitheca Avhich has been removed, and the coral- 

 lites are left intact and uninjured, -with their calices quite empty. 

 In other cases, tlie epitheca has been entirely decorticated, whilst the 

 corallites may remain uninjured, or may be more or less broken away 

 towai'ds their outer ends. Such specimens can in general be readily 

 recognised by the general shape of the colony and the peculiar 

 characters of the corallites. In other cases, lastly, the epitheca is 

 sufficiently thick to render the calices of the corallites below obscure 

 or invisible. In these instances, concentric lines of growth are 

 usually exhibited by the epitheca, and these are sometimes developed 

 into such strong and regular annulations as to similate pretty closely 

 the appearance of perfect specimens of CUsiophyllwm Oneidaense, 

 Billings. 



The corallites radiate from the imaginary axis of the colony, either 

 in straight lines or curves ; and the size of the mass in the turbinate 

 specimens increases rapidly by the interstitial addition of fresh coral- 

 lites. In shape the corallites are rounded, sub-prismatic, or more 

 commonly distinctly prismatic. In size they are by no means uniform, 

 there being generally a considerable number of under-sized corallites 

 intercalated amongst the nearly equal-sized larger tubes. The larger 

 corallites have most commonly a diameter of from a line to a tenth 

 of an inch, whilst the smaller ones may be half a line or less in width. 



The tabulae are commonly complete, sometimes incomplete, and 

 are about three or four in the space of a line. 



The mural pores, so far as I have observed, are uniformly in single 

 rows, placed on the flat surfaces of the corallites, not surrounded by 

 an elevated border, and of comparatively large size. Their distance 

 apart is most commonly about half a line, but is sometimes as much 

 as a line. 



The walls of the corallites are of unusual thickness, in the great 



majority of cases ; and they are not undistinguishably fused with 



those of contiguous corallites. Hence the lines of division between 



the walls of neighbouring tubes can be plainly seen in parts from 



4 



