526 T. A. STEPHENSON 
It has been so generally recognized that smooth-bodied and 
verrucose forms, and forms with and without acrorhagi cannot 
be separated into different families, that it seems hardly 
necessary to discuss this here. 
Secondly, there is the question of a separate Boloceridae. 
Such a fanuly has been in use by some authors, and originally 
I felt a need for it (see 1918 a, p. 19), but further work has 
changed that feeling. It hardly seems that the deciduous 
tentacles are a character giving the Boloceras any right 
to separation, and otherwise they are exactly Actiniidae. ‘This 
is especially the case since Boloceroides and Buno- 
deopsis have also the deciduous tentacles, and neither of 
them could be included in a Boloceridae in any case. One 
has to think of the cases as convergences. Evenif Bolocera 
and Bunodeopsis should be further stages, along different 
lines, from a Boloceroides-like ancestor, this is no reason 
for classing the three together. 
Thirdly, the Bunodidae. It seems a pity to have to attack 
an old-established family like this, but at the same time there 
seems to be no valid way whatever of separating it from the 
Actinidae (in the revised sense), with which it is continuous. 
Originally the Bunodidae relied fer separation upon their 
verrucae and their strong circumscribed sphincters. The 
verrucal character was swept away by Epiactis and Iso- 
tealia, which are without it. We must now tackle the 
sphincter. In the first place the sphincter in Bunodactis 
(Bunodes) itself is variable, and often not a strong one. In 
the type-species, B. gemmacea, it may be half diffuse in 
some cases (I have sections of a very typical specimen showing 
this—see ‘T'ext-fig. 11, p), and poorly developed. It is in 
Tealia and Epiactis (Text-fig. 12, a, B, c) that the really 
strong sphincters are found, and even there the size varies with 
species and individual. Further (this will be dealt with again 
under Bunodactis in Part II), there are apparently no 
criteria by which Bunodactis can be separated from 
Anthopleura and Actinioides, even generically—the 
three run right into each other and really form one large genus 
