76 ECHINODERMATA. 



slender, extending about an inch and a half, including the tentacula, and 

 rather under the eighth of an inch thick. 



In a few days, this specimen appeared at its full stretch on the side 

 of the vessel. The nniddle became gradually smaller and smaller, and 

 quite smooth, so that the figure of the animal resembled a long sand- 

 glass. In the afternoon of the day when this was observed, it had sun- 

 dered into two halves of nearly equal dimensions, each about half an 

 inch long, and when noticed, about four hues apart. The division 

 seemed to be natural, not the result of constraint, nor attended with 

 suftering. But in respect to the fti.sus, there is much presumption of 

 nmtilation following violence, though unobserved. 



The fact presumed, receives confirmation by attending to the course 

 of the process. 



Of five specimens obtained on August 11, four were dingy white, 

 tending to brownish, and one pure white. Next evening, two of the for- 

 mer were much extended cylindrically, the one two inches and a quarter, 

 the other exhibiting a contraction in the middle, two inches and a half 

 In expectation of division, I watched it some time, but unsuccessfully, 

 for this did not then take place. However, it ensued in the night. An- 

 other had divided also, and a third was observed next morning under 

 the process of division. 



In this third subject, the two halves were about five lines apart, 

 though still connected by a small greenish thread or ligament, apparently 

 tubular, which I conjectured might be the intestinal canal. The liga- 

 ment had subsisted above twelve hours. After its ruptui-e, the parts 

 Ijecame somewhat approximated, just as if they had mutually strained 

 themselves to be free. The process is represented in its advance on 

 August 13, fig. 22. 



The tentacula of four of the five specimens having been displayed 

 immediately preceding the ruptures undergone, it cannot be supposed 

 that the animals were then sufiering, as on the slightest constraint, these 

 organs are instantly contracted, or entirely withdra^vii. 



The upper portion of a divided specimen, fig. 15, as on July 19, is- 

 again represented as on August 5, by fig. 19. It had remained station- 



