BROWN, RUGOSE CORALS 
87 
17, to the typical Hapsipkyllum varsoviense adult condition which is shown 
in figure 19. This shows the completed inner wall, complete except for 
the gap at the cardinal septum. A slight change has also taken place in the 
condition of the tertiary septa. Instead of being short and close to the side 
of the primary and secondary septa they are now nearly or quite half as long 
as the secondary septa and are widely separated from them except at their 
inner margins. This condition is very clearly shown in figure 19. 
V. Observations on Development. 
From the foregoing studies of the Streptelasma group of the rugose corals 
it is seen that this group is first known from the true Streptelasma forms 
found in the middle Ordovicic. By gradual changes as we pass into the 
higher and geologically later horizons these very early forms give way to more 
complex and more specialized forms. The changes and specialization are 
not confined and directed along any one particular course, but form rather 
a gradually branching series, the various branches being more or less parallel 
to one another but at the same time quite distinct. Each one of these par¬ 
ticular branches or lines of development reaches a certain climax and then 
rapidly declines and disappears. At the close of the Devonic period only 
a few are left and at the end of the Carbonic all are gone, in North Amer¬ 
ica at least. 
The earliest representatives of the rugose corals yet recognized, namely 
the Streptelasmas of the middle Ordovicic, are by no means primitive 
although they are without a doubt ancestral to a large majority of the suc¬ 
ceeding genera and species. Streptelasma profundum is a well developed 
coral, although it is the earliest one recorded. The questions which now con¬ 
fronts us are: What was the ancestor of Streptelasma profundum ? and What 
was it like? This ancestral form has not yet been found, and until it is 
found we shall have to be content with deducing from all the known facts the 
probable characters of this ancestral form. In such deductions two differ¬ 
ent lines of investigation must be followed out; first a study of the early em¬ 
bryonic development of similar forms living at the present time; second a 
study of the early stages of development of all the available later forms. 
Since there are no living representatives of the rugose corals, and since the 
mode of development of the hexaeorals is so widely different from that of 
the rugose corals, we have to pass the first named line of investigation after 
drawing only a few of the most general inferences. 
Duerden has carried on extensive studies on the embryology of certain 
recent corals but in each case finds that the embryo is well started before the 
