28 PULSATION OF JKLLYFISHES. 



by cutting them out after the more complicated circuits had been set 

 into pulsation. Instead of simplifying the pulsating labyrinth, we 

 may increase its complexity, but as long as the waves proceeding from 

 the center can find a single uninterrupted circuit, the figure pulsates. 

 Thus, a disk cut as in figure 28, A, is set into pulsation and then all of 

 the inner rings are cut so as to be converted into "cut-off" paths as 

 in figure 28, B ; but the disk continues to pulsate until we cut across 

 the outermost ring, when it stops instantly. Every one of the forms 

 shown in figures 8 to iga can be thus stopped by even the smallest 

 cut which breaks the last circuit, although they continue to pulsate 

 despite any cutting which does not sever the circuit. Thus, figure 16 

 stops at once if we cut across one of the narrow places between the 

 rays of the star. 



The center of pulsation usually establishes itself in a large uncut 

 area, but once it be established we may greatly cut down this area 

 and not interfere with the center. Thus, the ring shown in figure iga 

 may be thinned by cutting at S, but the center remains undisturbed. 



Fig. 30. Very elongate circuits showing that the peripheral parts are better conductors of pul- 

 sation than are the inner parts of he sub-umbrella. These circuits can be caused to pulsate 

 continuously. 



Sustained pulsation without marginal sense-organs can be main- 

 tained only in tissue forming a closed circuit. These circuits may be 

 complex and constricted at intervals to mere thread-like connectives, 

 as in figures 31, A-c, where every annulus is crossed by radial cuts ; or 

 they may be very simple, as in figure 31, D. The circuits may either 

 cross or trend with the muscle fibers.* 



On two occasions disks were set into sustained pulsation when only 

 the marginal sense-organs were cut away ; no other cuts having been 



*The statement in my preliminary paper in the Carnegie Institution Year Book for 

 1905 that the circuits must trend with the muscle fibers is erroneous. 



