CLASSIFICATION OF ACTINIARIA BOT 
The first assumption, of independent acquisition, would not 
be unreasonable, but at the same time it does seem likelier 
that the ancestor of all three had ciliated tracts, perhaps only 
in a slightly differentiated form; and it is a simpler way of 
putting things to think of some forms losing them than of 
Text-ric. 17. ; 
Y 
t] 
shi, 
pe 
Mesenterial filaments. A and 8, a Zoanthid, with powerful ciliated 
tracts (f). E passes twice through these, as it cuts through 
a curved edge of mesentery. B, Edwardsia. Ciliated tracts 
present but less marked than in the Zoanthid; here and in D 
there are also reticular tracts (r). c, Paradiscosoma. Here 
there are no ciliated tracts, but three large sting-cells are 
shown. bD, Artemidactis. Typical Actinian filament, 
with median cnido-glandular tract (c) and lateral ciliated and 
reticular tracts. 
three groups gaining them. ‘There seems no special reason 
why such an ancestral form as that under consideration should 
not have weak ciliated tracts, because although very distinct 
structures they would easily be differentiated early on, just 
as acontia seem to have been at the Hosagartia stage m 
NO, 260 Pp 
