40 I'he Aiiierican Geologise. juiy, i89ft 
As is well known, colonies of Heliopora are dimorphic, consisting of 
siphonopores and autopores, the former produced from one another by 
intermural gemmation, the buds springing from the canals by which 
they are united. The latter are produced l)y the fusion of a group of 
siphonopores through the peculiar jirocess known as "coenenchymal 
gemmation." The autopores are tabvilated at long distances and pro- 
vided with pseudosepta; the siphonopores are aseptate and more closely 
tabulate. The pseudosepta range in number in mature cells from 11-10, 
in the same cells at different stages of growth from 2-16. The skeletal 
substance is composed of radially disposed, jjarallel, prismatic rods, 
which are generally placed each at the junction of four siphonopores 
and are, therefore, essentially quadrilateral. They are firmly united 
along their opposed sides by dental sutures, and they are excavated 
along their angles by the siphonopores (Nicholson, Man. Pal. vol. 1, pp. 
332, .333). These spicules in cross section show a simple or comyjound 
central axis and a radial structure. They are of about the same diam- 
eter as the siphonopores, perhaps a little larger. 
In Heliolites also the colony is dimorphic. The autopores are provided 
with septa, the number of which is almost invariably twelve in each 
corallite. Usually they are lamellar and often "extend to a considerable 
distance into the interior of the visceral chamber, sometimes meeting to- 
form a reticvilated columella {H. intrica), or even showing alternately 
large and small septa." (Nicholson, p. 337.) They are. therefore, true 
septa, and the symmetry of the polyp must have been hexameral, not 
octameral as in the Alcyonaria. The skeleton is not spicular and the 
walls are thin (in strong contrast to those of Heliopoi-a in every way), 
but those of the autopores are stronger than those between siphono- 
pores. The siphonopores reproduce usually by fission, seldom by inter- 
mural gemmation. The autopores, according to Sardeson, increase by 
coenenchymal gemmation, very much as in Heliopora, i. e. the several 
siphonojjores become confluent, the incipient autopore gradually in- 
creases in size, at the same time assuming thicker bounding walls and 
developing the regular number and arrangement of the septa. On the 
other hand. Nicholson describes the process as fundamentally different. 
The future autopore is first outlined by a thickening of the bounding 
wall, the included siphonopores still retaining their own proper walls. 
Then they are "seen to be suddenly arrested in their growth and com- 
monly to be cut off by a common tabula, their place vertically being 
taken by a single autopore" (Nicholson, p. .337). Nicholson's reputation 
as an accurate observer requires that Mr. Sardeson 's observation should 
be verified. 
In fact it appears that Heliolites resembles Heliopora principally in 
but two points; in being dimorphic, and in having coenenchymal gem- 
mation. On the other hand, it differs from Heliopora entirely in the 
skeletal structure of the corallum, in being provided with true septa, in 
having hexameral instead of octameral symmetry. Dimorphism i tself 
is a diagnostic feature of no great importance ; it is found in the fistu'i- 
porids, in many Bryozoa (Paljeozoic), even in FavoHites{f) canadensis . 
