27 
continuous with those of both series of septa. In this case there may be a dark line 
continuous for a short distance along the middle of the common zone; but its place 
may be taken by a row of dark spots, in which case the wall appears to consist of 
septa alone” 
Smyth corrects Nicholson’s (1888, 151) impression that the mural 
pores were tortuous and shows them to be straight. However, they are 
much rarer than in Calapoecia and the wall apparently is not rendered 
cribriform by their presence. There are twenty-eight septa, which as in 
Calapoecia occur as ridges; but they are not raised to spines as in that 
genus. Furthermore, they are continuous with the central reticulate tissue 
so characteristic of Cleistopora. Smyth also shows that tabulse are pres- 
ent in Cleistopora geometrica; their presence was unknown before his work. 
The interest of Cleistopora geometrica in this connexion lies in the 
fact that it is a colonial form with the intercorallite “ wall ” made up of 
septal elements alone; and also in that pores break this wall. But it differs 
from Calapoecia particularly in the presence of a loose, reticulate structure 
in the visceral cavities of corallites, in the comparatively feeble development 
of the pores, and in the number, and to a less extent, the habit, of the 
septa. Nicholson’s figure la-d and Smyth’s Plate VIII, figures 1-10, 
illustrate these points. ^ 
Dr. Bassler has kindly sent me specimens of Nyctopora and Lichenaria 
to compare with Calapoecia. Of these Nyctopora crenulata Bassler, from 
the Trenton (Hermitage) of Gordonsville, Tenn., comes nearest to the 
genus. It differs in having true corallite walls, which are not cribriform. 
The septa are lamellar and apparently not serrate. In the character of its 
septa N. crenulata is nearer to Calapoecia than to Favosites. 
It has been pointed out above that Calapoecia and Lyopora are not 
very dissimilar. In Lyopora the corallite wall is apparently made up of 
septal elements back to back. Although descriptions hitherto have 
emphasized the compactness of the wall in this genus, the present writer 
is of the opinion that such compactness may largely be due to preservation 
and that originally the wall might have been porose. This view is based 
on a study of thin sections (*See especially Plate II, figure 10). 
The genus Calapoecia has usually been regarded as one of the Tabu- 
lata, with the provision, in most cases, that C. anticostiensis is not a Cala- 
poecia (Nicholson 1879 and 1889; Zittel 1879, 1903, 1915, 1927; Roemer 
1883), It has been shown above that C. anticostiensis is unquestionably a 
Calapoecia, and, therefore, cannot be omitted in a discussion of systematic 
position. Sardeson (1896, 276) discusses a form which he calls “ II ough- 
tonia cf. huronica Rominger ” from the “? Upper Silurian ” of Cumberland 
sound (= Cumberland gulf, southeast Baffin island).- He describes this in 
^ I sent Professor Louis Smyth sections of G. canadensis for examination, and he has 
since replied (personal communication, April 1935). “There is certainly a resemblance 
in septal structure between Calapoecia and Cleistopora .... but the diflferencea (you 
enumerate them) seem to me to outweigh this likeness”. 
2 1 have never seen a reference to fossiliferons strata in this locality and it is 
possible that Sardeson is mistaken. This is supported by Soper’s (1923, 4) remark that 
“The rocks . . all about Cumberland sound are Pre-Cambrian.” Fossiliferous Ordovician 
rocks are known at the head of Frobisher bay, the large gulf immediately south of Cum- 
berland sound. Sardeson does not say how he acquired his specimen, but there was a 
German International Circumpolar Station at Sirmilling bay (Cumberland gulf) in 1882 
and it is possible that it reached him through a member of this. If it did, there is 
some reason for the mistaken locality. I have noticed, too, that the Eskimo often pick 
up and keep fossil corals if the structure is sufficiently obvious; it is even possible that 
the specimen reached Sardeson indirectly in this W’ay. 
