THE AXIAL GRADIENTS IN HYDROZOA. 1 93 



though it is possible that some portions may be resorbed, but 

 younger developing hydranths or buds may be completely re- 

 sorbed without visible loss of tissue. In Bougainvillea even the 

 younger medusa buds underwent complete resorption under the 

 more extreme inhibiting conditions, e.g., KNC w/ioooo six days, 

 then gradual decrease in concentration. After two weeks the 

 youngest medusa buds were completely resorbed, only the empty 

 perisarc marking their position, somewhat older buds were partly 

 resorbed and the most advanced buds had undergone disintegra- 

 tion. 



The question of the nature of the process of resorption of hy- 

 dranths was discussed some twenty years ago by Loeb ('oo) and 

 Thacher ('03), Loeb postulating a liquefying enzyme and Thacher 

 interpreting the process as a degeneration. To all appearances 

 the process is to some extent a real dedifferentiation. There is 

 no visible loss of tissue and from what we now know of metabolic 

 relations in the hydroids a simple interpretation of resorption 

 and retraction of parts seems possible. The growing hydranths 

 and medusa buds are regions of higher metabolic rate than the 

 stems and are therefore able to appropriate the larger share of 

 nutrition, but at the same time they are more susceptible to in- 

 hibiting or depressing conditions than stems (Child, '19, '21). 

 When subjected to these conditions their metabolic rate decreases 

 to a much greater degree than that of stems. Under such condi- 

 tions they may not only be unable to maintain themselves, but 

 their tissues, instead of taking nutriment from other parts, may 

 become, whether through autolysis or other factors, a source of 

 food for other parts. Consequently they may undergo decrease 

 in size until nutritive equilibrium is established, or untilthe 

 region with low r er metabolic rate is more or less completely re- 

 sorbed. Whether individual cells actually die in such a process 

 or merely undergo dedifferentiation is difficult to determine, but in 

 the light of what we know of the possibilities of dedifferentiation 

 in these animals, it seems probable that cell death does not neces- 

 sarily occur. 



In starving hydroid stocks in the laboratory the retraction of 

 ccenosarc in some regions and its outgrowth in others has often 

 been noted. Tests of susceptibility and of permanganate re- 

 duction in such stocks indicate that the regions which are grow- 



