CLASSIFICATION OF ACTINIARIA 565 
the Actiniidae, took to a floating life, swimming upside down. 
The base developed into a regular float, and certain anatomical 
peculiarities appeared. These are the Minyadidae. 
Other advanced stages are represented by the various 
families of *‘ Stichodactylines ’. These arose from some Actiniid 
or pre-Actinid ancestors, and they have usually the numerous 
perfect mesenteries, and often endodermal sphincters, which 
may be quite strong. The sphincter is endodermal or absent, 
never mesogloeal. Among themselves they diverged into seven 
families, easily distinguished from one another. The differ- 
ences affected the arrangement of the tentacles on the endocoels 
and exocoels, and their form—they might be simple, pinnate, 
dendritic, sessile and vesicular, feathery, modified into special 
stinging ‘ nematospheres ’, and.so on. The other part of the 
structure chiefly affected by variation was the musculature— 
there might be absence of sphincter in one case compensated 
for by strong retractors ; or very strong sphincter and retractors 
but poor tentacles ; and so on. 
It will be seen from the above outline, and from that given 
earlier for Mesomyaria, that there is one thing assumed as 
having independently taken place in Endomyaria, Mesomyaria, 
and Endocoelactaria—and possibly more than once in Endo- 
myaria: that is, that in each of these cases a start was made 
from the condition in which the mesenteries are divided into 
macro- and microcnemes, this was lost, and in the end there 
were graded mesenteries and numerous perfect pairs. This is, 
however, a convergence quite to be expected among forms 
making in a general way towards increase of size and diameter 
of the individual, and correlated multiplication of organs. 
Whatever arrangement be adopted, there is some convergence 
cropping up, but when one thinks of the vertebrate and 
cephalopod eye, or of the Marsupial and ordinary wolf, a 
cenvergence like that assumed here seems very simple. 
In the two hypothetical ancestors of Endomyaria and 
Mesomyaria (Hoactinia and Eosagartia) there is no 
harm in assuming for them ectodermal muscle in the body-wall, 
and the same may probably be said for the Hoactinia-like 
