50 THE ROYAL SOCIETY OF CANADA 



induction shocks has the same effect as a series of stimuli so spaced 

 that one is thrown into the muscle just after the conclusion of the 

 absolute refractory period associated with the response to the last 

 stimulation. Thus the duration of the excited state of each re- 

 sponse becomes shorter and the rate of propagation becomes slower 

 until the characteristics essential for the circulating wave are attained. 

 Once started the existence of fibrillation tends to maintain .the con- 

 ditions necessary for its appearance. 



Yet under some conditions a stimulus of very brief duration may 

 induce fibrillation. I propose to describe here some experiments 

 showing how the relation of the time of application of such a brief 

 stimulus to the cardiac cycle may be of great importance in deter- 

 mining whether or no the stimulus will initiate fibrillation. 



The production of fibrillation in the cooled ventricles of the rabbit by the 

 application of a properly timed stimulus. 



In these experiments the hearts of rabbits were profused with 

 Ringer's solution (NaCl 0-9%, KC1 0-042%, CaCl 2 0-024%, NaHC0 3 

 0-1%) by Langendorff's method, using the perfusion apparatus of 

 Brodie. The perfusion was started at body temperature but was 

 allowed to fall gradually to room temperature. A thread attached 

 to the apex of the heart and to a lever enabled the ventricular move- 

 ments to be recorded on a kymograph. As a rule, non-polarisable 

 electrodes were placed in contact with an auricle and a ventricle and 

 connected to an Einthoven galvanometer so that records of the elec- 

 trical changes could be recorded when desired. A pair of platinum 

 electrodes, separated by two or three millimetres, rested against one 

 of the ventricles. These were connected with the secondary of an 

 induction coil. Stimuli were delivered by single taps of a Morse key 

 connected with the primary of the stimulating coil by means of a 

 double relay so arranged that the moment of application of the stimulus 

 was signalled by the simultaneous use of a sparking coil connected by 

 a short air-gap with the insulated metal pointer and to the drum. It 

 was found in a number of experiments that a single tap of the Morse 

 key if properly timed would start fibrillation which would persist for 

 a time which varied in different cases from a few seconds up to 

 over three hours, then giving place suddenly to a normal sequence of 

 beats. The fibrillation affected the ventricles only and was never 

 transmitted back to the auricles. 



The point of interest is that the stimulus employed would never 

 cause fibrillation unless it was set in at a certain critical instant. 

 Figures 3, 4, and 5 illustrate the phenomenon. In Fig. 3 the first 

 stimulus applied falls within the refractory phase and does not in- 



