264 Mr. S. Bidwell. On Changes produced by [Mar. 25, 



Addition, April 3rd, 1886. 



It would be difficult for anyone who has not actually seen the 

 apparatus above described to appreciate its extreme delicacy, and the 

 accuracy with which it is capable of measuring such minute quan- 

 tities as would commonly be regarded as infinitesimal.* It has been 

 suggested to me that greater value would be attached to the expe- 

 rimental results contained in the present and former papers if the 

 manner in which they were arrived at were described in greater 

 detail, and a few of the actual scale readings given in full. 



In the case of the twelve series of observations to which the table 

 given in this paper relates, my method of proceeding was as follows : — 

 The iron wire having been placed in position and loaded with a 

 weight, a short time was allowed for the apparatus to attain a nearly 

 steady temperature. The reflected image of the indicating wire 

 (which I will call the index) was theu, by means of the fine screw 

 adjustment, brought upon the upper half of the scale, the zero point 

 of which was in the middle, and the index, which, owing to small 

 variations of temperaturef, was rarely absolutely at rest, was watched 

 until its upper edge nearly coincided with one of the scale divisions. 

 The number of this division was noted and recorded from my dicta- 

 tion, and at the instant when exact coincidence occurred, a contact 

 key was depressed, which caused a current of 1*6 ampere to pass 

 through the coil. The number of the scale division nearest to which 

 the index was deflected was again noted and recorded as before ; and 

 if the point which the index reached happened to be exactly midway 

 between two divisions, the reading was recorded to half a scale divi- 

 sion. When the deflections were small, the readings were taken to 

 the nearest half scale division ; but this, though easy enough, was in 

 general considered to be a needless refinement. 



The next course was to find by a tentative method the strengths of 

 the three currents which respectively produced — (1) the first sensible 

 elongation ; (2) the maximum elongation ; (3) neither elongation 

 nor retraction ; and in order to do this, the resistance in the circuit 

 was varied by means of a large set of coils, and a succession of currents 

 of different strengths (perhaps from twenty to fifty in number, or 

 sometimes even more) were caused to pass through the apparatus. J 



* The instrument was exhibited in action at the Soiree of the Royal Society, May, 

 1885. 



f Taking the coefficient of expansion of iron to be 0*0000122 per degree C, the 

 heat elongation due to a rise of temperature of one degree would produce a deflec- 

 tion of sixty-one scale divisions ; but in addition to the iron there was a somewhat 

 greater length of brass, and if this shared in the heat expansion, a total deflection of 

 more than 150 scale divisions per degree would be produced. 



X It is of course understood that the circuit was actually closed by the key only 



