28 C. M. CHILD. 



protect this region from the action of external agents. Thi.-. 

 however, is not the chief factor in the low susceptibility and re- 

 ducing power of this region, for when these stages are scraped 

 from the substratum the attached surface is often left exposed, 

 but even then the original apico-basal gradient does not appear. 

 As a matter of fact, various experiments on susceptibility and 

 reducing power of naked protoplasmic surfaces and those covered 

 by perisarc have shown that, except when the perisarc is old and 

 very thick, it is rapidly penetrated by most agents. The very 

 thin perisarc of stages such as Figs. 16 and 17 is penetrated rapidly 

 by KMnO 4 and the agents used to determine susceptibility 



The tip of the growing stem which represents the high end 

 of the secondary gradient gives rise to the first hydranth (see 

 Fig. 8). This hydranth itself shows a marked apico-basal gradient 

 which is a part of the hydranth-stem gradient. Each tentacle, 

 originating as a localized region of growth on the hydranth body, 

 represents a gradient with the high end at the tip. The tentacle 

 gradient apparently originates in the localization of the region 

 of growth giving rise to the tentacle. This region is not sharply 

 defined, but its activity decreases peripherally from a central area 

 and this decrease evidently constitutes the beginning of the ten- 

 tacle gradient. The gradients of numerous other axiate organs 

 undoubtedly arise in the same way. The localization of a region 

 of growth giving rise to a tentacle must be determined by cor- 

 relative factors in the developing hydranth. The tentacles arise 

 at a certain level of the polar gradient and at a certain distance 

 from each other. Each growing tentacle region, once localized, 

 apparently dominates a certain area as does a growing plant bud 

 and determines the course of development of this area. From 

 this viewpoint the tentacle is a gradient in a specialized part which 

 originates at a certain body level in reaction to the polar gradient 

 of the body. In the species serving as material for this paper the 

 tentacles of a circle are apparently not simultaneously determined 

 or localized but the factors concerned in their localization are 

 obscure. 



