166 
MR. G. J. ROMANES ON THE LOCOMOTOR SYSTEM OF MEDUSiE. 
which the rest of the Medusa presented cannot have been due to any variations in the 
quality of the contact between the electrodes and this stationary seat of stimulation. 
(f) Any such merely mechanical source of fallacy being thus, I think, excluded, 
we are compelled to regard the facts of artificial rhythm as of a purely physiological 
kind. The question, therefore, as to the explanation of these facts becomes one of the 
highest interest, and the hypothesis which I have framed to answer it is as follows. 
Every time the tissue contracts it must as a consequence suffer a certain amount of 
exhaustion, and therefore must become slightly less sensitive to stimulation than it 
was before. After a time, however, the exhaustion will pass away, and the original 
degree of sensitiveness will thereupon return. Now the intensity of faradaic stimula¬ 
tion which is alone capable of producing rhythmic response, is either minimal, or 
but slightly more than minimal, in relation to the sensitiveness of the tissue when 
fresh. Consequently when the degree of this sensitiveness is somewhat lowered by 
temporary exhaustion, the intensity of the stimulation becomes somewhat less than 
minimal in relation to this lower degree of sensitiveness. The tissue, therefore, fails 
to perceive the presence of the stimulus, and consequently fails to respond. But so 
soon as the exhaustion is completely recovered from, so soon will the tissue again 
perceive the presence of the stimulus ; it will therefore again respond, again become 
ternporarily exhausted, again fail to perceive the presence of the stimulus, and again 
become temporarily quiescent. Now it is obvious that, if this process occurs once, it 
may occur an indefinite number of times ; and as the conditions of nutrition, as well 
as those of stimulation, remain constant, it is manifest that the responses may thus 
become periodic. 
(g.) In order to test the truth of this hypothesis, I made the following experiments. 
Having first noted the rate of the rhythm under faradaic stimulation of minimal 
intensity, without shifting the electrodes or altering the intensity of the current, I 
discarded the faradaic stimulation, and substituted for it single induction shocks 
thrown in with a key. I found, as I had hoped, that the maximum number of these 
single shocks which I could thus throw in in a given time, so as to \procure a response 
to every shock, corresponded with the number of contractions which the tissue had 
previously given during a similar interval of time when under the influence of the 
faradaic current of similar intensity. To make this quite clear I shall describe the 
whole course of one such experiment. The deganglionated tissue under the influence 
of minimal faradaic stimulation manifested a perfectly regular rhythm of thirty con¬ 
tractions per minute, or one contraction in every two seconds. While the position of 
the platinum electrodes and the intensity of the current remained unchanged, single 
induction shocks were now administered with a key at any intervals which might 
be desired. It was found that if these single induction stimuli were administered at 
regular intervals of two seconds or more, the tissue responded to every stimulus; 
but that if the stimuli were thrown in more rapidly than this, the tissue did not 
respond to every stimulus, but only to those that were separated from one another 
