49^ T. H. Morgan. 



first place the ligature does not in some cases entirely prevent the 

 beginning of the development of the oral end and occasionally 

 I have seen, as Stevens and I have previously observed, that the 

 primordium of the oral hydranth may be laid down. Further- 

 more, if the material in the circulation is derived mainly from the 

 broken-dovi^n ridges, etc., of the oral end it is not clear yv\\y a 

 similar breaking down might not also occur at the aboral end and 

 in this way by doubling the total amount present make possible 

 the simultaneous development of both hydranths. Other diffi- 

 culties are also present that may be spoken of later. In the hope 

 of gaining some further insight into these questions the following 

 experiments were carried out: 



The hydranths were removed from a number of pieces, the 

 oral ends of some of these pieces were tied at once (A); others 

 were tied at the end of six (B); or of twelve hours (C). Now 

 during the first six to twelve hours after cutting, the endodermal 

 ridges of the oral ends (in the pieces not yet tied) begin to break 

 down and their material is thrown into the circulation. If the 

 presence of this material in the circulation has any influence on 

 the rate of development of a hydranth (oral or aboral) it might 

 accelerate the aboral development if the oral end is now tied. 

 A number of experiments of this sort show that the aboral develop- 

 ment often takes place sooner in pieces whose oral end is tied after 

 SIX hours or often even after twelve hours than in check pieces tied 

 at once. Unfortunately the material began to "go bad" before a 

 sufficient number of results could be obtained to place the con- 

 clusion entirely beyond doubt, but there seemed to be evidence in 

 all cases of some acceleration in the development of the aboral 

 hydranth in pieces tied later than in those tied at once, and in 

 most cases the acceleration was so great that the former developed 

 even before the latter. Some of the more satisfactory experiments 

 may now be described. 



In the first experiment some pieces were tied at once (A); 

 others after six hours (B). Forty-seven hours later the A-pieces 

 showed nothing, while four of the seven B-pieces had produced 

 primordia. After seventy-two hours two of the A-pieces had 

 primordia, while four of the B-pieces had primordia, one a 

 hydranth, and two nothing. 



In another series, in which the B-pieces were tied after fourteen 

 hours, the aboral hydranths of the A-series developed first. The 



