IN THE STOMODiEUM OP THE ALCYONARIANS. 
699 
noglyphe in the polyps, whilst holding the view that the coenenchymal tubes represent 
siphonozooids in which the stomach, mesenteries, &c., have degenerated. 
The tendency of the dimorphic forms is, as I have pointed out, to throw the siplionic 
function upon the siphonozooids and to eliminate it from the autozooids. 
If, for any reason, it was of advantage to any dimorphic Alcyonarian to diminish 
the power of the circulation, this would be done by the gradual atrophy of the siphono- 
glyphe in the autozooids, and were this diminution insufficient the siphonozooids, or 
their siphonoglyphe alone, would become smaller and smaller. An example of this 
kind of process is presented by Renilla. In this genus there is but a small amount of 
fleshy sarcosoma, but there are large canal spaces which occupy the greater part of the 
colony, and here we find, owing probably to the need of only a feeble current, very 
small siphonozooids. In Pennatula, which presents a considerable quantity of sarcosoma, 
the siphonozooids are comparatively large. 
In Helioporct, on the view that the coenenchymal tubes represent siphonozooids, we. 
should have to suppose that the siphonozooids became smaller and smaller, then 
completely atrophied, and subsequently a siphonoglyphe re-appeared in the autozooids. 
This would obviously necessitate a stage in their history in which there was no 
siphonoglyphe, which would be a condition very difficult to understand. 
It is also difficult to believe that the mouth, stomach, and mesenteries would have all 
completely disappeared in this way, for even in the lowermost depths of the long body- 
cavities of such forms as Tubipora, Alcyonium, &c., two or more ridges remain, 
indicating the position of the mesenteries, and we should at least expect to find some 
such trace of the mesenteries in the degenerate siphonozooid. 
Remarks on the classification and pjhylogeny ofi the Alcyonaria. 
At present it can hardly be said that the classification of the Alcyonaria is in a 
satisfactory condition for two reasons, firstly, because no serious attempt has yet been 
made to trace the probable steps of the phylogeny of the group, and secondly because 
it is based on external zoological differences between genera rather than on the general 
features of their anatomy. 
Taking the classification in Claus’s ‘ Grunclzuge der Zoologie ’ as the one most 
generally adopted, we find such obvious incongruities as the following : the grouping- 
together of such colonial genera as Alcyonium, Clavularia, &c., with the simple 
isolated Haimea, Hartea, &c., the position of Paragorgia amongst the Gorgoniche, 
and so on. Recently, G. von Koch (13) has suggested a classification that is based 
on the varieties of the skeleton, but it seems to me that the Pennatulidse and 
Gorgonidae are not so closely related as to justify their position in the same division of 
the same group ( Axifera ). 
In presenting the following speculations on the phylogeny of the Alcyonaria, I am 
fully aware that the great difficulties of this group are only just beginning to be 
