THE PULSATION -STIMULUS IS NON-MUSCULAR. 21 



muscular tissue adjacent to it. There are a few spindle-shaped (gang- 

 lion?) cells upon the basal membrane at the base of the regenerated 

 epithelium, and occasionally one sees a large rounded cell in the 

 gelatinous substance below the basal membrane. Occasionally these 

 rounded cells have one or more delicate processes which extend into 

 the gelatinous substance. 



Figure C is a somewhat slanting section through regenerating sub- 

 umbrella tissue about 4 days old, which is beginning to regenerate the 

 muscle fibers and can now contract feebly. The muscle fibers appear 

 as elongate processes of deep-lying epithelial cells, and extend par- 

 allel one with another over the basal membrane, trending circumfer- 

 entially around the sub-umbrella. 



Figure D is a surface view of newly regenerated sub-umbrella epi- 

 thelium which transmits the pulsation-stimulus, but can not yet con- 

 tract, as it still lacks muscular elements. 



It is well known that Carlson has demonstrated the nervous nature 

 of the stimulus which produces pulsation in the heart of Limulus. 

 Indeed, I believe that all of the facts brought to light by Gaskell in 

 his attempt to prove the muscular nature of the transmission of the 

 stimulus of pulsation in the vertebrate heart will apply equally well if 

 we assume that the impulse is transmitted by diffuse nervous elements. 

 In the heart of the loggerhead turtle I find that the stimulus causing 

 pulsation is transmitted entirely through the thin outer muscular part 

 of the wall of the heart, and the thick cavernated inner part of the 

 heart's wall may be cut away without affecting the pulsation. Also, 

 the stimulus to pulsate is not transmitted through this cavernated 

 tissue to the muscular tissue. 



Fig. 5B. — Showing that the sub-um- 

 brella tissue is a better electrical 

 conductor than is the gelatinous 

 substance of the bell. The cur- 

 rent travels around through the 

 long way, rather than across the 

 shallow scratches which insulate 

 the area B. 



The sub-umbrella tissue of Cassiopea is a good conductor of electri- 

 city, while the gelatinous substance of the Medusa is a poor conductor. 

 Thus in fig. 5 b, if we insulate an annulus by the shallowest possible 

 scratch through the sub-umbrella, and then isolate a small sector, 

 B, by shallow radial cuts ; on touching the large sector A at 1 and 2 



