290 



HARRY BEAL TORREY. 



produced may, for the time, not detain us) in the original physio- 

 logical system. In another respect they differ, in that they lead, 

 in the one case, to a hydranth, or part of a hydranth, in the other 

 to a holdfast. This difference is essentially an expression of the 

 different conditions controlling their development, of which the 

 influence of adjacent parts is the chief. The intimacy of this 

 coordination is obviously a function of the physiological isolation 



of the parts concerned. It is a 

 conspicuous fact that Y figures 

 are formed almost invariably 

 from short heteromorphic seg- 

 ments; shrunken, starving, slow 

 developing (Fig. 13, which shows 

 eth beginning of a bud) pieces give an especially large proportion 

 of them. Frequently that portion of a segment of the column 

 which lies against the floor of the aquarium puts forth tentacles 

 more slowly than the upper surface. This retardation in develop- 



FIG. 13. 



FIG. 14. 



ment, due probably to diminished supply of oxygen next the sub- 

 stratum, is accompanied by a bending of the segment by means of 

 a contraction of the affected side. It is on the opposite, convex, 

 aspect of the column that the bud develops (Fig. 14, in which 

 tentacles and gonads are only partly drawn), not, therefore, in 

 direct contact with the substratum. And it is significant that, 

 with the rarest exceptions that are referable to exceptional con- 



