48 EEPOKT — 1888. 



diameter through which the incandescent surface could be observed. 

 This arrangement was placed in circuit with eight or ten cells of a 

 secondary battery, an ammeter and an adjustable resistance being included 

 in the circuit. It was necessary that the resistance should be capable of 

 being taken out very slowly, so that the current might be gradually and 

 regularly increased. 



The arrangement devised answered very well in practice. It consisted 

 of a frame across which was stretched a series of lengths of germao 

 silver wire, some of them being straight and others coiled so as to give 

 varying amounts of resistance in each length. By means of plugs any or- 

 all of the lengths could be put in and out at pleasure, and by this means the- 

 resistance could be roughly adjusted. The fine adjustment was given by 

 a cross-piece working along two of the straight wires and moved by a 

 moderately fine screw. As the cross-pieces moved along the length of 

 the wires it, by connecting the two, brought more or less resistance into 

 the circuit. The resistance of the whole arrangement amounted to about 

 one ohm. It was found that with this arrangement the strip of platinum 

 could be rapidly raised to incandescence, and could be kept close to the- 

 melting-point for a considerable time, a very slight increase in the- 

 current being then sufficient to cause fusion. The photometric tests- 

 made with this arrangement confirmed those which had been made when 

 the platinum was fused by the oxyhydrogen jet ; that is to say, the 

 observations showed considerable irregularity, so much indeed that it 

 did not seem worth while to make, as had originally been intended, a long 

 series of them, nor to construct apparatus more accurate than the first 

 experimental one. 



A decided defect in the apparatus was the buckling of the platinum. 

 As it was tightly gripped at each end there was no room for expansion,, 

 and before melting there was considerable expansion and consequent 

 buckling. Had it seemed worth while to do so, it would not have been, 

 difficult to devise an apparatus by which this might have been obviated, 

 but the results were not sufficiently encouraging. 



A few photographic tests were also made by permitting the light from 

 the incandescent platinum to fall through a screen with openmgs in it 

 upon a sensitive plate, a number of exposures from diS'erent pieces of 

 platinum being made on the same plate. These of course could only be 

 looked upon as rough tests, but they also seemed to indicate that the 

 amount of radiation from a given surface of platinum at the moment of 

 fusion is not absolutely constant under the conditions we have described, 

 and so far as they are worth anything they may be taken as confirming the 

 conclusion obtained by the other methods. 



