and Urticating Cells. 255 
tain adhesive granules remained adherent to any object which 
could not be drawn nearer, and then, during the contraction 
of the guard-animal, the thickened terminal part was much 
drawn out in length. If then another part of it adhered to 
some other object and the machopolyp contracted still further, 
it sometimes acquired, in a certain sense, a ramified form. 
These guard-polyps are exclusively offensive in their 
action. If the polyp-stock be disturbed they retract them- 
selves into their chitinous calices. I can completely confirm 
Allman’s statement*, that the “ pseudopodium-like processes” 
are visible only on pertectly fresh stocks, as regards the adhe- 
sive polyps; and I believe that the statements of Allman and 
Hincks relate principally to these adhesive polyps. We shall 
recur to their function further on. 
3. Guard-Animals with Urticating Capsules and Adhesive 
Cells. 
These machopolyps are found exclusively in the genus 
Aglaophenia; and here they occur in those nematophores 
which are placed behind the nutritive animals and project 
beyond these. I call them paired superior guard-animals. 
For our investigation some species of Aglaophenia with 
large superior guard-animals which are abundant here (Mel- 
bourne) are admirably adapted. The nematophores which 
contain these machopolyps have, besides the circular orifice 
at the extremity, also an oval aperture at the side; this is 
placed towards the nutritive animal, at the spot where the 
chitinous calice of the nutritive animal terminates, on the 
concave anterior side of the nematophore which is turned 
towards the nutritive polyp. In these nematophores are 
seated guard-polyps which are composed of two parts. In 
the distal extremity, close under the terminal orifice, there are 
urticating capsules; while from the proximal aperture a fila- 
ment can be extended, at the extremity of which there is 
an adhesive knob. ‘The distal part resembles an animal armed 
with urticating capsules, while the proximal part is to be 
compared to an adhesive animal. 
Of the former, however, only that part is developed which 
represents the distal part of an animal armed with urticating 
capsules, such as 1 have described above. ‘There is no endo- 
dermal cellular axis, and therefore also no supporting lamella. 
But ganglion cells occur in the subepithelial layer; and be- 
tween the cnidoblasts there are radial muscles and supporting 
* Loe. Cty Pp: LID. 
