46 THE ROYAL SOCIETY OF CANADA 



spread it out on a glass plate, pour on serum and cover up with a 

 vaselined watch-glass. Preparation remains quiescent for quarter 

 of an hour. Pricking with a needle point provokes a strong contraction. 

 Wave runs round ring in each direction; the waves meet on the op- 

 posite side of the ring and die out. Repeated the stimulus at diminish- 

 ing intervals and after several attempts started a wave in one direction 

 and not in the other. The wave ran all the way round the ring and 

 then continued to circulate going round about twice a second. After 

 this had continued for two minutes extra stimuli were thrown in. 

 After several attempts the wave was stopped. The preparation then 

 remained at rest for ten minutes. The circulating excitation was 

 again started in the same way as before. This time there was con- 

 siderable difficulty in stopping the wave. A number of attempts 

 caused slowing of the wave in its passage over part of the course, but 

 failed to arrest it. Presently a single stimulus was so timed as to 

 arrest the wave. The preparation then remained absolutely quies- 

 cent. There was no sign of 'automatic' rhythm throughout this 

 experiment. After cutting through the ring in one place, the strip 

 of tissue responded with a single contraction to each of a series of 

 stimuli if suitably spaced. It was found impossible to get more than 

 one response to a single stimulus. 



I have repeated the experiment successfully on five preparations 

 from dog-fish auricles. A large heart must be taken otherwise it is 

 difficult to secure that the duration of the refractory state shall be 

 shorter than the time taken by the wave to pass round the ring. The 

 chief error to be guarded against is that of mistaking a series of auto- 

 matic beats originating in one point in the ring and travelling round it 

 in one direction only owing to a complete block close to the point of 

 origin of the rhythm on one side of this point. The cleanest experi- 

 ments are those, such as that quoted above, in which the auricle 

 showed no tendency whatever to give spontaneous beats. Severance 

 of the ring at that point will obviously prevent the possibility of cir- 

 culating excitations but will not upset the course of a series of rhyth- 

 mic spontaneous excitations unless by a rare chance the section 

 should pass through the point actually initiating the spontaneous 

 rhythm. 



Ordinary graphic records either mechanical or electrical are of 

 no value in attesting the occurrence of a true circulating excitation 

 in rings of this kind, since the records show merely a rhythmic series 

 of waves and do not discriminate between a spontaneous series of 

 beats and a wave of excitation which continues to circulate because 

 it always finds excitable tissue ahead of it. The only method of 

 recording the phenonenon which I have found of any use is cinemato- 



