178 MR. G. J. ROMANES ON THE LOCOMOTOR SYSTEM OF MEDUSAE. 
phenomenon, however, it is necessary that I should detail all the facts I have been 
able to collect. I have never seen any similar or corresponding action performed by 
the bells of other medusids, and even in the case of Sarsia its occurrence is com¬ 
paratively rare. When it does occur, however, it is always continuous ; that is to say, 
it never spontaneously recommences after having once ceased. As already stated, the 
period of its duration is extremely variable; but when this period is long, it is observ¬ 
able that the shivering motions become feebler and feebler, until they eventually fade 
away into quiescence. The animal is then quite dead to all further stimulation. 
Beyond saying that the peculiar motions in question never originate independently of 
stimulation, I cannot give much further account of the conditions which determine 
their commencement. The following instances are quoted from my notes in extenso: 
£ A healthy individual with centres removed, after failing to respond to either make or 
break of direct current, and after about a quarter of a minute’s rest from a series of 
rapidly alternating makes and breaks of this current, began to shiver, and continued 
to do so for five minutes. Afterwards quite dead.’ 
“ ‘ Another healthy individual, after refusing to contract either on make or break of 
induced current for two or three times, began to shiver, and continued to do so for 
twenty minutes. Afterwards quite dead. This individual had been used five minutes 
before for experiments with the direct current, to which it responded well and without 
shivering. ’ 
“ ‘ Another healthy individual was left for some time after excision of margin, and 
then put into the well and submitted to induction-shocks. No contraction either on 
make or break with coil pushed to zero. On now trying direct current, without 
altering position of electrodes, violent contraction on make and also a decided one on 
break. On again trying induced current, no contraction. On again trying direct 
current, strong contractions [thus far, of course, the behavior of the tissue was normal, 
see above § 2 (A)], and after four or five of these, shivering began ; this lasted for four 
minutes, and, when again quiescent, the bell again responded to make and break of 
direct current in the ordinary way.’ 
“ Now all that can be gathered from these and similar notes is, that the shivering 
motion in question may be started either by the direct or by the induced current, and 
this in some cases when the bell has repeatedly refused to answer stimuli in the 
ordinary way. It may further be gathered from these notes that shivering is most 
likely to begin after the bell has received a number of shocks in succession.” 
The passage concludes by observing that the shivering movements hi question are 
much more certain to be produced by means of chemical stimulation, as just described 
in the previous paragraphs of the present paper. Now, when we consider these 
results of electrical stimulation in the light which has since been shed by the results 
of all the other experiments on artificial rhythm, there can be no doubt that in some 
way or other the injury which the electrical stimulation has previously supplied to the 
tissue continues for a long time afterwards to act as a constant stimulus to these 
