''Polarity." 499 



start of fourteen hours was so great that even the acceleration of 

 the B-pieces did not suffice to make them develop first, and this 

 would hardly be expected since the whole development often 

 occurred in less than forty-eight hours, but the B-pieces were not 

 fourteen hours behind the A-pieces. 



In another series tied at once (A); after six hours (B), and 

 after eighteen hours (C), it was found after forty-eight hours that 

 five of the A-pieces barely showed primordia and five others 

 nothing; that four of the B-pieces showed primordia further 

 advanced than the primordia of the A-pieces, and six nothing; 

 that one of the C-pieces showed the barest beginning of a primor- 

 dium and nine nothing (in poor condition). After seventy-two 

 hours all of the A-pieces had primordia; four of the B-pieces had 

 hydranths and the remaining six primordia; all ten of the C- 

 pieces had young primordia. 



In an experiment of this kind different pieces are necessarily 

 compared, and, since no two pieces can be assumed to be cut at 

 exactly equivalent levels, it is, perhaps, unsafe to draw conclu- 

 sions from so small a number of observations. If, as seems to be 

 the case, pieces tied after six hours develop as soon as, or sooner 

 than, those tied at once (/. ^., six hours earlier), the result is 

 probably due to the presence of material in the circulation derived 

 from the oral end before tying. 



It may be asked, why may not the reserve material of the w^all 

 throughout the piece be used rather than that thrown into the 

 circulation by the breaking dow^n of the ridges .f* If this were the 

 case the total amount of material would be much more than neces- 

 sary to supply both ends of a long piece at once, as shown by the 

 fact that a long piece cut into shorter ones will produce as many 

 oral hydranths as there are pieces, in the same time that long; pieces 

 cut at equivalent levels produce oral hydranths. Therefore if an 

 appeal is made to the amount of food material to explain the 

 acceleration of the aboral hydranth, it must be the material of the 

 circulation postulated rather than that of the wall. 



In still another way I have tried to find out what part, if any, 

 of the materials of the circulation influence the rate of develop- 

 ment. The hydranths were cut off from a number of pieces and 

 then after six hours (or more) ligatures were tied around the pieces 

 at different levels, in some pieces (A) near the oral ends; in others 

 (B) near the middle; and in others (C) quite near the basal end. 



