84 



Dr. A. D. Waller. An Attempt to Estimate 



My attention at this early stage of the inquiry has been chiefly 

 directed to the deterioration of seeds with age and to the comparison 

 inter se of sets of seeds of certificated years by means of the germina- 

 tion test and of the blaze test used quantitatively. 



I selected beans as being of suitable bulk and readily obtainable, and 

 I have to thank Messrs. Sutton for supplying me with many different 

 samples of known dates. After a considerable number of trials upon 

 entire seeds variously orientated between the electrodes, soaked in 

 water of various temperatures for various periods, and upon the several 

 isolated parts of seeds, I fixed upon the following procedure as con- 

 veniently yielding series of numerical results comparable inter se. 



The " dry " beans are first soaked in water for twelve hours in an 

 incubator adjusted at 25° C, then laid upon moist flannel and replaced 

 in the incubator for examination during the next day. Each bean was 

 then peeled and split, and the radicle was carefully broken off and placed 

 between the clay pads of the electrodes (fig. 1) so that the uninjured 

 apex was in contact with the upper electrode A, and the fractured 

 base with the lower electrode B. With this position we have a 

 "'positive " current of injury from B to A, and have to expect a " nega- 

 tive blaze " current from A to B in response to excitation. In order 

 that the response shall be " unequivocal," the exciting current is taken 

 of negative direction. To ensure maximal effect a strong current is 

 taken, viz., a break induction shock at 10,000 units of Berne coil. 

 And inasmuch as a current of such strength repeated for a second time 

 shortly after a first trial produces little or no effect, and even when 

 repeated after a considerable interval a much smaller effect than at its 

 first application, it is necessary to take for the purpose of numerical 

 comparison exclusively the values obtained at first trials. To this end 

 it may be necessary to shunt the galvanometer to such an extent that 

 the blaze effect to be expected from the first excitation shall give a 

 deflection within the scale; a second trial when the first trial has given 

 a deflection off scale, is of no value whatever. 



By adoption of uniform conditions on these lines, comparisons may 

 profitably be made between different series of results. But at this 

 early stage of the inquiry, not knowing what conditions it might be 

 advisable to select, I have been forced to vary them in tentative direc- 

 tions, by variation of strength of excitation,* of length of soakage, and 



* To avoid exhaustion by strong currents, and to obtain a regularly repeated 

 series of effects, I find that condenser discbarges are more suitable tban induc- 

 tion shocks. The discharge of 1 microfarad charged by two Leclanche cells (= about 

 40 ergs) usually gives a convenient normal effect upon which to investigate the 

 effects of temperature variations, and of anaesthetic vapours. 



I also find it preferable to use the radicle some hours after it has been broken off, 

 by which time its current of injury has subsided, and blaze currents are obtainable 

 in both directions. 



