288 
We find f. inst. that the hydrothecae of Lafoza. fruticosa and L. 
gracillima show a distinet adcauline convexity. As to the other 
sertularian characters the genera Perisiphonia, Cryptolaria and Gram- 
maria have -their hydrothecae adnate to the-hydrocaulus in a larger 
or smaller pårt of their adcauline wall, and in the two last named 
genera this connection is as in the Sertulariidae inseparable, while 
in Perisiphonia it can be loosened by the aid of reagents. Ås, 
therefore, the above named forms cannot by a single character be 
divided from the Lafoéidae, I, on the contrary, find it arbitrary 
to refer them to the Sertulariidae, and at the same time there 
can be no doubt, but that the latter family gains in firmness 
and coherence by only embracing operculate forms. 
In animals of so simple å structure as the Hydroid polyps we 
can only expect to find a few distinguishing marks between the 
systematic divisions, and most families and genera are only divided 
from each other by one or two characters. The two chief divisions, 
the Athecata and the T'hecaphora are only divided from each other 
" by the presence in the latter of more or less developed proteetive 
cases for the hydrants and the gonophores, and it seems reasonable 
to ascribe systematie significance also to the operculum, a structure, 
which must be regarded as the complement of the proteetive cases 
Jand, so to speak, as the end-result of the same effort, which has' led 
to the formation of the hydrothecae and gonothecae. We may 
"further point out as an evidence of its systematie importance, that 
it has that in common' with other structures of systematie signifi- 
cance that it presents a rich development of characteristic net 
vcations which give excellent generic characters. 
We may now regard the relation between the Laftkåns and 
the Campanulinidae. If we compare the two above family diag- 
noses we shall find that the two families are only sharply divided 
by a single character, namely the presence ør absence of an; oper- 
"eulum, all: the" other charåcters being more or less relative, and, 
therefore, there can be no dende, that they are rer MAG. related 
