EEPOET ON THE HYDROIDA. 
xiii 
the general structure of the ectoderm in the Hydroida. If, however, Jickeli be correct 
in assigning a nervous function to certain cells which he finds generally distributed in 
the ectoderm of the Hydroida, and which he regards as ganglion cells (see below, p. xvii), 
then the physiological significance which Kleinenberg attributes to his neuro-muscle-cells 
must be somewhat modified. 
While in the Hyclroid trophosome the muscular fibrillte of the ectoderm are thus 
always longitudinal it is different witli the Medusae. Here a circular musculature is 
largely and characteristically developed in the ectoderm of the sub -umbrella or concave 
side of the walls of the bell-cavity, while the velum is itself mainly an offset of the 
ectoderm in which circularly directed fibrillse constitute the chief part of its substance. 
While the muscular fibrillse of the trophosome, whether in the ectoderm or in the 
endoderm, show no trace of striae, the fibrillae which run in a circular direction in the 
umbrella of the Medusa are transversely striated. 
The ectoderm in most Hydroids is separated from the surrounding perisarcal tube 
by a wider or narrower space. Across this space, whether it be in the stems or in the 
gonangia, numerous narrow processes are sent off from the outer surface of the ectoderm 
to become attached by their ends to the inner surface of the perisarc. In these processes 
no distinct cell boundaries can be detected, and the same is very frequently the case in 
the ectoderm from which they proceed. In the living animal a most interesting proto- 
plasmic movement may be seen in them, the processes constantly changing their form, 
frequently throwing out branches like the pseudopodia of certain Rhizopods, and ex- 
tending and withdrawing themselves across the intervening space. When they reach 
the perisarc their ends flatten themselves out on it and emit radiating filaments of proto- 
plasm. From these flattened ends the rest of the process frequently breaks away, 
leaving the extremity with radiating protoplasmic filaments attached to the walls where 
they have exactly the appearance of typical stellate cells (PI. XXV. figs, la, 2b). 
Among the Hydroids of the present Report are species (Pis. XXVII. fig. la ; 
PI. XXXIII. fig. 2a) in which similar ectoderm bands are seen stretching from the body 
of the hydranth to the walls of the hydrotheca, a condition which I have also observed 
in some British Sertularians. In the Challenger species in which those bands occur 
the hydranth would seem to be incapable of complete retraction. 
Reichert 1 was the first to call attention to the emission of pseudopodia-like filaments 
by the ectoderm, and was led by it into the erroneous belief that the Hydroid coenosarc 
consists of an undifferentiated protoplasm, — a view which, not obtaining the acceptance 
of any zoologist who had studied the structure of these animals, resulted in his very 
interesting and significant observation of protoplasmic movements in the coenosarcal 
processes remaining long unrecognised. 
Hamann 2 has shown that a similar emission of pseudopodial filaments takes place 
1 Carl B. Reichert, Tiber die contractile Substanz und ihre Bewegungs-Erscheinungen, Berlin, 1867. 
2 0. Hamann, loc. cit. 
