16 
Dalarne, Sweden. There were no figures of this species. The original 
description (Lindstrdm, 1880, 36) is insufficient to distinguish a form from 
C. canadensis var. anticostiensis. From the remarks made above it wdll be 
remembered that the remoteness of corallites is a variable feature as is their 
size. “ Coenenchyma vesiculosum ” is sufficient to invalidate the compari- 
son with Lyopora favosa Nich. and Eth. as it is understood at present. 
Since C. aniphigenia Lindstrom cannot be distinguished in its original 
description from C. canadensis var. anticostiensis, and because examples 
from the type locality referred to this species prove to be referable to the 
same variety, it is proposed to include C. arnphigenia Lindstrom in C. 
canademis var. anticostiensis. Troedsson (1928, 127) is of the opinion that 
C. arnphigenia should be included in C. borealis Whitfield. It will be shown 
below that the species C. borealis Whitfield is not recognizable and also 
that Troedsson's conception of that species is based on an arbitrary selec- 
tion of characters, all of which are found in the rather variable var, anti- 
costiensis. Lindstrom’s measurement of corallite diameter w’as probably 
external, not internal, as given in this paper. 
The specimen (R.M.S. Ca. 602) is a small colony with a few' calices 
appearing at the surface. These are obscurely seen but appear to be more 
remote, for the most part, than is the case in the type of var. anticostiensis. 
Sections contradict this appearance, however, for in them the corallites are 
in fairly close proximity. Unfortunately no longitudinal section can be 
obtained from this specimen. A section (N.R.S. Ca. 604) of a specimen 
from Boda, Dalarne, shows the same pink preservation as Ca. 602; it has 
corallites in close proximity. The section, which is transverse, does not 
cut any pores, hence the inner end of the costae is obscured. 
Two specimens (Ca. 569, 570) from Losa Sternar, Visby, Gotland, can 
readily be identified with var. anticostiensis. In sections both these speci- 
mens are identical with the Dalarne material, and hence are supposedly 
C. arnphigenia Lindstrom.” As in the latter case, the remoteness of cor- 
allites as seen at the surface is belied by the sections, where in many cases 
they appear in close proximity, some even in contact. The coenenchyme 
is comparable with the less regular of the Anticosti specimens and shows 
a good development of the curved elements between the horizontal dia- 
phragms. The “ costse ” are, perhaps, shorter than in the Anticosti speci- 
mens. 
The reasons for assigning Houghtonia cf. huronica of Sardeson 1876 
to this species are given in the discussion of Relationships of Calapoecia on 
a later page. 
Calapoecia canadensis var. anticostiensis forma arctica Troedsson 
(Plate IV, figures la, b, 2a, b, 3) 
Sarcinula organum Lam. Etheridge, 1878, 586-7. 
Calapoecia arctica Troedsson 1928, 127-9; PI, 39, PI. 40, figs, la, b. 
Diagnosis. Calapoecia canadensis var. anticostiensis with large coral- 
lites and peripheral incomplete tabulse slightly more numerous than is 
typical in that variety. 
The present writer has cut and examined a piece of the holotype of 
C. arctica Troedsson (M.M.K. 3027), Plate IV, figure 2, and another speci- 
