25 
shortness, and the character of the wall, which is thicker and much less 
cribriform (i.e. the pores are smaller). The tabulae, too, are more remote 
and more regular than in forma arctica, although we have seen this is a 
variable character in Calapoecia. The pores occur opposite the space 
between each of the principal diaphragms, that is about five in 6 mm. 
The other examples of this species show the horizontal diaphragms 
of the coenenchyme becoming progressively more remote; but pairs remain 
in close proximity so that the spacing out of their diaphragms is effected 
by a zone of non-deposition of coenenchyme for a certain distance above 
each pair. A specimen from the “Lower Silurian, Lepiaena-limestone; 
road from Boda to Osmundsberget, Riittoik district ” Dalecarlia (B.M. 
R30150), shows a rather irregular arrangement of diaphragms which are 
more remote than in the examples first described. The “ costse '' can be 
seen as low ridges on the external surface of the corallites, but where the 
diaphragms meet this surface they extend out on to them and, just as in 
Calapoecia, occupy a polygonal area around the corallite. These polygonal 
areas are seen to perfection in one of M’Coy's specimens (S.M. A6654) 
from the Coniston limestone (Bala) of Coniston Water. It is the negative 
impression of the upper surface of S. organum and shows, moreover, that 
the polygonal areas round each corallite are in contact. The rows of pores 
in B.M. R30150 are 1-5 mm. apart, which is considerably more than in 
the first-described examples. Similar to this is a specimen collected by 
G. J. Hinde from Gotland and kindly lent to me by Dr. Stanley Smith. 
It differs in that the diaphragms are laid down in a concave fashion and 
turn up sharply before they reach the corallite wall, with the result that 
in this region each successive diaphragm lies over that immediately suc- 
ceeding it as an open cone over cone. In these examples the diaphragms 
are not arranged in pairs. 
Another specimen of M’Coy's, S.M. A6657 from the Coniston lime- 
stone of High Haume, England, shows the diaphragms definitely arranged 
in pairs which are fairly remote from the next (Plate IV, figure 7). The 
corallites communicate with the space contained within the diaphragm 
pairs through oval pores, which lie on horizontal rows. The rows of pores 
are about 3 mm. apart. An example from Yamlingbo, Gotland (B.M. 
25852), is similar to this, but the mural pores are about 2 mm. apart (Plate 
IV, figures 4a, b). These two specimens are linked by one collected by 
Professor A. E. Trueman from the Upper Llandovery of Oslo and kindly 
lent to me. Here the pores vary from 2 to 3 mm. apart. Rows of pores 
3 to 5 mm. distant are seen in B.M. 56147, from Group 4, Oslo. Here the 
diaphragms lie comparatively straight. 
In these few specimens alone there is a sufficiently gradual variation to 
show that they cannot be satisfactorily placed in separate groups distin- 
guished by the structure of coenenchyme. In this morphological series we 
have traced a gradual rearrangement of the simple type of coenenchyme 
identical with that seen in the most developed Calapoecias. It is possible 
that the squamulse described by Lindstrom (1882, 72, and Taf. VI, 5 and 
6), and seen in all but the most Calapoecia-like of the specimens mentioned 
above, represent vestigial diaphragms. Troedsson (1928, 120, 123) has 
already remarked on this, but he did not notice the morphological relation 
