MR, G. J. ROMANES ON THE LOCOMOTOR SYSTEM OF MEDUSAE. 
167 
by an interval of at least two seconds’ duration. Thus, for instance, if the shocks 
were thrown in at the rate of one a second, the tissue only, but always, responded to 
every alternate shock. And similarly, as just stated, if any number of shocks were 
thrown in, the tissue only responded once in every two seconds. Now, as this rate of 
response precisely coincided with the rate of rhythm previously shown by the same 
tissue under the influence of faradaic stimulation of the same intensity, the experiment 
tended to verify the hypothesis which it was designed to test. 
(h.) But in order to test this hypothesis still more effectually, I conducted a number 
of experiments on a slightly different plan. As already stated, the intensity of the 
faradaic stimulation has a marked influence on the rate of the rhythm, up to the 
point at winch the rhythmic effect of such stimulation begins to become lost in the 
tetanic effect. Well, by choosing at random any strength of faradaic stimulation 
between the limits where rhythmic response occurred, and by noting the rate of the 
rhythm under that strength of stimulation, I found that I was generally able to 
predict the precise number of single induction-shocks which I could afterwards afford 
to throw in with the same strength of current, so as to procure a response to every 
shock—this number, of course, corresponding with the rate of the rhythm previously 
manifested under the faradaic stimulation. I say “generally,’’ because this experiment 
was not invariably successful—any more, I may add, than was the first-mentioned and 
almost identical experiment in which minimal stimulation was employed (see para¬ 
graph g). Nevertheless, the exceptions were but slight, and always on the side of 
somewhat fewer contractions occurring in a given time in response to the single shocks 
than in response to faradaic stimulation. And as this is the direction in which we 
should expect, from the principle of the summation of stimuli, that slight deviations 
from the ordinary rule would occur, if they occurred at all, I do not think that the 
fact of then- occurrence tends to impair the confirmation which these experiments 
certainly afford to the hypothesis we are considering. 
(i.) Indeed, it seems to me that this hypothesis is so fully substantiated by these 
experiments, that I feel it is ahnost superfluous to adduce another experiment having 
the same tendency. Nevertheless, as the facts which the hypothesis is intended to 
explain are facts of considerable importance in themselves, and as the hypothesis in 
question will be shortly employed as tending to further the theory of ganglionic action 
in general, it is desirable that I should state all the experiments by which I have 
hitherto endeavoured to test it. 
The experiment which I am about to describe cannot be understood without a 
preliminary account of certain properties of the contractile tissues of Aurelia, which 
as yet I have not described. The properties in question all appear to arise from the 
following fact—namely, that the contractile tissue admits of great variations in the 
vigour of its contractions. In my last paper, while treating of the summation of 
stimuli, it was fully explained that the force of responsive contraction in A urelia varies 
greatly in the different parts of a “staircase” series; but quite apart from this 
