EXPERIMENTS ON MED US& 29 



confined to two small regions (a and b, Fig. 6) which 

 correspond to the sinus venosus and the bulbus aortse 

 of the frog's heart. When we excise these two pieces 

 from the heart they continue to beat without inter- 

 ruption, while the long part between the two pieces 

 no longer pulsates (in sea-water at least). These ex- 

 periments, it seems to me, leave no room for doubt 

 that the change in the direction of the contraction in 

 the Ascidian's heart is determined by each of the two 

 ends getting the upper hand alternately, and forcing 

 the other centre to act in its rhythm for a time. This 

 " getting the upper hand " might possibly mean no- 

 thing more than that one end gains the time in which 

 to send off a wave of contraction before the other 

 end begins to contract. For this it is only necessary 

 that a single heart-beat of the leading end be delayed 

 or fail entirely, a phenomenon that also appears oc- 

 casionally in the human heart. In this way the other 

 end of the heart gains time in which to send out a 

 wave of contraction, and its automatic activity will 

 continue to be the stimulus for the activity of the 

 first end until a delay occurs in one beat or until one 

 beat is skipped, thus allowing the first end time again 

 to become automatically active, and so on. 



Last year I asked the members of the class in gen- 

 eral physiology at Wood's Holl to find out whether 

 the latter view was correct. Their observations were 

 as follows : Suppose at a certain time a to be the 

 active and b the passive end of the heart. After a 

 short time a begins to beat more slowly or ceases to 



