IO 



HARCITT. 



not a constant feature in normal specimens, in many of which 

 gonads are lacking on one or more canals. Occasionally the 

 fusion of the cut edges and the formation of new canals resulted 

 in apparent abnormalities. Fig. 1 1 shows such a case in which 

 the new canal a has anastomosed with one of the old ones. 

 But this again has its 'counterpart among normal specimens in a 

 state of nature. 



The single quadrants which were cut out in the preceding 

 experiment were placed in separate dishes. These contained a 

 single canal, in some cases a part of a manubrium. The contrac- 

 tion of the cut edges took place as usual and in from three to 

 four days the typical medusa-form had been assumed, the edges 



FIG. 10. 



FIG. ii. 



of the margin and marginal canal had fused as had also the 

 velum. The apex of the quadrant was flexed toward the mar- 

 gin and when the fusion was complete this portion, and with it the 

 proximal end of the canal, was joined to or very near, the mar- 

 ginal canal. In specimens having a portion of the old manu- 

 brium this was also carried downward and fused with the marginal 

 canal and a gastric pouch formed at the point of union, and when 

 the new manubrium was finally regenerated it protruded beyond 

 the edge of the velum in the same horizontal plane. In speci- 

 mens where no portion of the old manubrium was present usu- 

 ally entirely new ones regenerated but in the same general posi- 

 tion as in the former. These new medusas were of course only 

 about one fourth the size of the original specimen and with only 

 a proportional number of tentacles. 



