14 
used merely to facilitate the making of this rather important point; of 
course it is not a structure, but only the result of the distribution and 
habit of the septal elements. These remarks are also applicable to longi- 
tudinal sections, where the pores will show a shape depending entirely 
on wnere the cut is made. 
The tabulae are both complete and incomplete. A central count gives 
ten to twelve in 5 mm. They may pass through pores into the coenenchyme, 
but are typically bent upward at the periphery of the corallite, crossing 
the face of a pore and joining one septal spine to that above it. These 
upturned tabulae may give the false appearance of a wall in a transverse 
section cut through a region of pores. 
The coenenchyme is well developed and consists primarily of hori- 
zontal elements which correspond fairly well with the number of tabulae 
in adjoining corallites and may commonly be continuous with these. Here 
and there curved plates run from one horizontal diaphragm to the next 
in a way exactly like the upturning of the tabulae at the periphery of the 
corallites. The “ costae ” run into the coenenchyme and subdivide it verti- 
cally. The coenenchyme varies from regular and simple to a rather untidy 
arrangement within the single corallum. 
The significance of this structure and its relation to that of C. 
canadensis is discussed on later pages {Variations in Calapoecia) , 
This specimen is labelled “ West side of Gamache bay, Anticosti 
island; Division 1, Anticosti group; collected by T. C. Weston.” Its 
horizon, then, following Schuchert and Twenhofel (1910) would be Ellis 
Bay (Gamachian) of the Cincinnatic. Twenhofel (1928, 130) compared 
specimens collected from Anticosti island with the type of C. anticostiemis 
Bill, and found them identical. But it has been suggested above that some 
of these are probably C. canadensis; Twenhofel evidently considered both 
species the same. Twenhofel describes longitudinal keels both on the 
inside and outside of the walls. . . . The number varies from 18 to 25 and 
in nearly, if not all, cases, they extend about 0-5 mm. beyond the corallite 
walls.” These keels evidently are a conception that includes all the septal 
elements, ridges, spines, and “ costae.” These have already been considered 
in detail from thin sections and hence an analysis of Twenhofel’s concep- 
tion is unnecessary. But the writer must repeat that never has he seen 
more than twenty septa, although he has been careful to check this as 
often as possible in sections. The several descriptions of the septal parts 
of Calapoecia have led to confusion in the past. For instance Etheridge 
(1903, 18-19) was puzzled by the apparently conflicting accounts of Lambe 
(1899) and Nicholson. The former writes of ‘Spiniform septal ridges' 
and figures a section (Plate I, figure 6) where the septa appeared as spines 
projecting from a definite wall; Nicholson refers to “longitudinal ” ridges. 
This apparent contradiction was due to Lambe, who did not make it clear 
that the “ wall ” was made up of septal ridges from which spines projected. 
In addition to the type, the writer has sectioned and examined speci- 
mens from the G. J. Hinde collection (B.M. R21535, R21534, R21514, 
R21513) (Plate III, figure 6) . These were collected from “ Near West 
End (i.e. West cape), Anticosti, Canada. Anticosti group, Division B, Cin- 
cinnati formation,” which is the horizon of the Vaureal (Twenhofel 1928). 
They are identical with the type, but the orderly arrangement of the coenen- 
