ON THE OCCURRENCE OF THE GENUS COLUMNARIA — ETHERIDGE. 31 



for about one-third of their diameter, or perhaps a little more, whilst 

 the latter are mere marginal crenulations. These primary septa 

 are tapering and spike-like in cross-section, although at the same 

 time true lamellae, extending from top to bottom of the corallites. 

 There is not the faintest trace of any meeting of the septa in the 

 calice centres to form a spurious columella as described in C. rigida, 

 Billings.* At first sight the corallites appear to be provided with 

 very few septa in consequence of the small size of the secondary, 

 and even these are set far apart. Furthermore, the septa do not 

 spring sharply from the corallite walls, but in consequence of the 

 inner edges of the latter being concave between them, a more or 

 less festoon-like appearance is given to the cross-section of each 

 corallite, somewhat as one sees in the genus Ueliolites. These 

 appearances at first caused doubt to arise in my mind as to the 

 propriety of referring this coral to Columnaria, but on referring 

 to Prof. AUeyne Kicliolson's excellent figures,! I found that in both 

 C. aheolata and C. calicina very much the same features existed. 



The stereoplasmic thickening of the septa is unequal, some 

 being thin and spike-like ; others, from a greater preponderance 

 of this deposit, becoming club-shaped (PI. viii.. Fig. 6). Tn some 

 corallites the secondary septa become scarcely, if at all developed, 

 in others they assume the character previously described. 



In no instance have I noticed an undue predominance in length 

 of one or more septa, a point in which C. pauciseptata differs from 

 C. calicina at least, but there is certainly no regular development 

 of four septa as in Stauria, nor tiie slightest trace of a division 

 into cycles. Many of the corallites are partially infilled with a 

 dendritic growth of iron oxide fringing the septa. 



On longitudinal weathered surfaces, the primary septa appear 

 as strong continuous lamellae, their paucity and larger comparative 

 size rendering them conspicuous objects. 



The tabulae (PI. viii.. Fig. 7) are particularly well developed, 

 simple, complete, mostly horizontal, very seldom thickened, opposite 

 in contiguous tubes, or very slightly alternating, in other words 

 sub-opposite. They vary from three-quarters to one millimetre 

 apart, and in a few rare instances are somewhat more distant from 

 one another. The non-horizontal are simply bent or curved in 

 some portion of their course, never vesicular or incomplete. The 

 diaphragm forming the floor of the calice is striated by the septa 

 passingon to it. The intertabular or old visceral chambers are nearly 

 square, from the fact of the transverse measurements of the tubes 

 and the distance apart of the tabulae being so nearly coincident. 



There is not the slightest trace of the existence of the mural 

 pores, or intramural canals, so characteristic of the Favositidae. 



* Teste Nicholson, loc. cit., p. 196. 

 t Loc. cit., t. 10, f . 1 and 2. 



