the Vitality of Seeds by an Electrical Method. 



85 



of interval between soakage and examination. These departures from 

 strict uniformity, while affording necessary information, restrict legiti- 

 mate comparisons to data within each particular table ; comparisons 

 from table to table may not be safely made. 



Fig. 3. 



o 5 /o /5 ao 25 somm. 



Photographic record of an unequivocal blaze current of the radicle of a bean 

 (1900 crop). Excitation by a strong break induction shock in the A to B or 

 negative direction. Hoinodrome response of 0"075 volt. 



With regard to the germination tests, they have been carried ouc for 

 the most part upon similar lots taken from the same parcels as those 

 from which other seeds were taken to be electrically tested as described 

 above. This latter required each seed to be broken up and rendered 

 unfit for germination. I think that the parallel pair of tests made 

 upon twin lots of different individual seeds is nearly as conclusive as if 

 both tests had been made upon the same individual seeds — vide, e.g., 

 Table I. Nevertheless, to meet the criticism that this proof is not 

 conclusive, I have obtained three series of data in which the electrical 

 and germination tests were carried out upon the same individual beans. 

 In all three series I previously determined the coefficient of each intact 

 seed by the blaze test ; the germination test was subsequently carried 

 out in one series at Kew under the supervision of Sir W. Thiselton- 

 Dyer (Table VII) ; in a second series at Chelsea under the supervision 

 of Professor Farmer (Table VIII) ; and in the third series by myself 

 in my own laboratory (Table IX). But I find it far less satisfactory to 



