706 
ME. G. J. ROMANES ON THE LOCOMOTOR SYSTEM OF MEDUSAE. 
any part of the mutilated nectocalyx is irritated, the polypite shows a very marked 
tendency to touch some point in the tissue-mass a a d e (fig. 8) by which it still remains 
in connexion with the bell, and through which, therefore, the stimulus must pass in 
order to reach the polypite. And it is observable that this tendency is particularly well 
marked if the section has been planned as represented in fig. 8 — i. e. in such a way as 
to leave the tissue-tract aade pervaded by a nutrient tube d e, this tube being thus left 
intact. When this is done the polypite most usually points to the uninjured nutrient- 
tube d e every time any part of the tissue-band a l is irritated*. 
( g ) Let us now very briefly consider the inferences to which these results would seem 
to point. The fact that the localizing power of the polypite is completely destroyed as 
regards all parts of the bell lying beyond an incision in the latter conclusively proves, as 
already stated, that all parts of the bell are pervaded by radial lines of differentiated 
tissue, which have at least for one of their functions the conveying of impressions to the 
polypite f. The fact in question also proves that the particular effect which is produced 
on the polypite by stimulating any one of these lines cannot be so produced by stimu- 
lating any of the other lines. Or, to state the same thing in the words previously used, 
we have conclusive proof that, so far as their function is concerned, these radial lines of 
differentiated tissue are undistinguishable from what we elsewhere know as afferent 
nerves. But although these tracts of differentiated tissue thus far resemble afferent 
nerves in their function, we soon see that in one important particular they differ widely 
from such nerves ; for we have seen that, after they have been divided, stimulation of 
their peripheral parts still continues to be transmitted to their central parts, as shown 
by the non-localizing movements of the polypite. Of course this transmission cannot 
take place through the divided tissue-tracts themselves ; and hence the only hypothesis 
we can frame to account for the fact of its occurrence is that which would suppose these 
tissue-tracts, or afferent lines, to be capable of vicarious action. Such vicarious action 
would probably be effected by means of intercommunicating lines, the directions of which 
would probably be various. In this way we arrive at the hypothesis of the whole con- 
tractile sheet being pervaded by an intimate plexus of functionally differentiated tissue, 
the constituent elements of which are capable of vicarious action in a high degree (see 
severity of section in fig. 8). And this hypothesis, besides being recommended by the 
consideration of its being the only feasible one that is open, is confirmed by the fact that 
the stimuli which it supposes to escape from a severed phalanx of radial lines, and thus 
* It may here he stated that the greater sensitiveness of the nutrient tubes is also made apparent by the 
responses of the nectocalyx ; for to a nip or prick of the general contractile sheet the response by the nectocalyx 
usually consists, as before stated, in a local bending in of the part irritated ; whereas in response to similar 
irritation of a radial tube, the response is usually a general bending in of the whole nectocalyx — i. e. a general 
spasm. 
I may here observe that cutting through all the nutrient tubes at their insertion in the polypite has no effect 
in preventing the localizing action of the latter as regards other parts of the bell. 
t We must not, however, conclude that these lines are radial structurally ; the evidence only proves that 
they are so functionally. 
