196 C. M. CHILD. 



occurs gradually and only after the stolon attains considerable 

 length, but in Obelia it usually occurs at a relatively early stage 

 and the region of separation is more sharply localized (Figs. 15, 

 17). Such differences, however, are not entirely constant for the 

 species, for Bougainvillea stolons sometimes show a definite level 

 of separation and Obelia stolons sometimes do not. 



After separation of the coenosarc the stolon may remain con- 

 nected with the stock by the perisarc. In the case of stolons 

 hanging free in the water the empty perisarc usually breaks and 

 the stolon falls to the bottom, attaches itself and continues to 

 grow, the tip growing at the expense of more basal levels until 

 exhaustion occurs, or until conditions are altered so that the tip 

 can transform into a hydranth and stem. Such free stolons may 

 cover many centimeters of distance, leaving behind them a tube 

 of empty perisarc as they go, and decreasing in length as their 

 substance is gradually used as nutrition. In the laboratory this 

 growth may continue for three weeks or even more, according 

 to temperature, and while transformation into hydranths often 

 occurs in the early stages, it apparently does not take place, 

 even in favorable environment, in the later stages, but the stolon 

 continues to "creep" over the bottom until reduced to a minute 

 amount of cellular material. And even when growth ceases the 

 small masses of tissue in the perisarc remain alive for some time 

 longer. 



The separation and continued growth of these stolons receives 

 a simple physiological interpretation in terms of the axial gradient. 

 If the stolon is such a gradient, the levels of relatively high rate 

 are able to live to some extent at the expense of lower levels. 

 Under laboratory conditions, without intake of food, the growth 

 of the stolon tip is possible only at the expense of other parts. In 

 the early stages the stolon tip, as a region of higher metabolic 

 rate than the old stem ccenosarc, is able to take material from the 

 latter, but as the stolon elongates the growth of the tip occurs 

 more and more exclusively at the expense of the lower stolon 

 levels, because the stolon gradient, and consequently the nutri- 

 tive concentration gradient, is limited in length and when the 

 length of the stolon exceeds this limit, it can no longer draw on 

 the stock for nutrition. 



From this stage on, the lower levels of the stolon gradient 



