166 J. PLAYFAIR MCMURRICH 
same two species, at the stage in which the acontia occur, the 
mesenteries of the fourth couple are longer than any of the others 
that are present, and the same relations may be seen in other 
forms described by Van Beneden, as for instance, in species of 
the genera Dactylactis and Ovactis, and in forms occurring in the 
Siboga collection. On the other hand, in those forms in which 
the mesentereis of the second couple are the most ventral ones 
provided with acontia, as in the genus Apiactis as described by 
Van Beneden and as I have found in the genus Peponactis, it is 
these same mesenteries that are the longest. 
Bearing in mind the fact that in the genus Pachycerianthus it 
is the fourth couple of mesenteries that become the telocnemes, 
it seems fairly certain that in those larvae in which the mesenteries 
of the fourth couple are the longest and at the same time the most 
ventral acontiferous ones, these mesenteries become the teloc- 
nemes. Or to make the statement more general, it is probable 
that the most ventral acontiferous mesenteries become the tel- 
ocnemes. It may be supposed, therefore, that in Arachnactis 
albida and A. sibogse the telocnemes are formed by the fourth 
couple of mesenteries, and this genus is, therefore, to be associated 
with Pachycerianthus. But the peculiarity is not confined to 
these two genera, occurring also in the larval genera Dactylactis 
and Ovactis, and therefore it has, apparently, more than generic 
value. It seems to me advisable to divide the acontiferous Ceri- 
anthese into two families, according as the telocnemes are the 
second or the fourth couple of mesenteries, and for the one family 
I would suggest the name Cerianthidse, while the other may be 
termed the Arachnactidae, this latter family including the genera 
Arachnactis, Dactylactis, Ovactis and Pachycerianthus. 
No stages of A. brachiolata and A. bournei have yet been stud- 
ied that are sufficiently advanced to show acontia, and while 
the mesenteries of the second couple decidedly surpass those of 
the fourth couple in length in the oldest known larvae, yet this 
may be merely a growth condition, the latter couple not yet 
having had time to acquire their future length relations. How- 
ever, even if the probability that in these species the telocnemes 
are the second couple be disregarded, their dissimilarity from 
