282 
The four families may be defined in the following manner: 
Campanulariidae : The campanulate or cup-shaped hydrothecae 
which not rarely present longitudinal ridges and marginal teeth 
are as a rule provided witb a well-developed stalk, and with few 
exceptions ( Hypantliea , Silicularia ) show a poly-symmetiical stiuctuie. 
A diaphragm is always present, but sometimes only in the shape 
of a marginal thickening. Nematophores are never present. No 
operculum. [A club-shaped proboscis]. 
Lafoeidae : The elongate, cup-shaped, retort-shaped or tubular, 
short-stalked or sessil hydrothecae, which never present longitudinal 
ridges or marginal teeth, not rarely show a more or less pionounced 
bilateral symmetry, and their adcauline wall is in man) cases in 
different extension more or less firmly connected with the stems 
and branches. A diaphragm may be absent or developed in different 
degree. Nematophores are sometimes present. No operculum. [A 
conical proboscis]. 
Campanulinidae : The more or less elongate, cup-shaped, 
pitcher-shaped or tubular, sessile or stalked hydrothecae, which 
never present longitudinal ridges or marginal teeth 1 ), sometimes 
show a bilateral symmetry, and are rarely provided with a com- 
plete diaphragm. Nematophores are sometimes present. A differently 
formed operculum. [A conical proboscis]. 
Sertulariidae : The pitcher-shaped, retort-formed or tubulai, 
bilaterally developed hydrothecae lack a free stalk, and the ad- 
.—— - - - - > 
i) S. F Clark (16, p., 12) has described a species of Campanulina, C. 
denticulata , in the hydrothecae of which he thought lie found be- 
sides an operculum of convergent segments an equal number of large 
castellated marginal teeth arising outside the latter. As both the 
marginal teeth and the operculum must have been formed as a 
secretion from the ectoderm we cannot understand the formation of 
the teeth, as the corresponding part of the ectoderm should have 
been placed outside the operculum. There can be no doubt but that 
the “castellated teeth” are only the worn segments of an old oper¬ 
culum. The figures 6 a and 6 b, therefore, represent two hydrothecae, 
which have been regenerated three times, and the only hydrotheca 
(fig. 7) in which the author could “tind no trace of teeth, only an 
operculum” is a hydrotheca which has not been regenerated. 
