212 Proceedings of Eoyal Society of Edinhurgh. [sess. 
4418 ohms at 25'". Thus there was a decrease of resistance to the 
amount of 0’019 ohms, or *43 per cent. 
Another wire, of diameter J mm. and of the same length as 
before, was put into the hath for the same length of time. The 
original weight was 1*5460 grammes, and the final weight T5494 
grammes ; so that, under the same assumptions, the volume of the 
charge of hydrogen was 38 cubic centimetres, or about 200 times 
that of the wire. The resistance decreased from 0*732 ohms to 
0*728 ohms, ^^e., by 0*55 per cent. 
Since a wire of diameter J mm. weighs 1*546 grammes, a wire of 
equal length of diameter i mm. will weigh *247 grammes, provided 
they are of the same material. But the one experimented with 
weighed only *217 grammes, and increased by 2*2 milligrammes. 
Hence the thicker one would have increased its weight under the 
same treatment by 2*5 milligrammes. In the experiment the 
thicker wire increased its weight under an approximately similar 
treatment by 3*4 milligrammes. Hence the thicker wire absorbed 
more hydrogen than the thinner wire in the ratio of 34 to 25, or 
1*36 to 1. Simultaneously, the resistance of the thicker wire 
diminished more than did that of the thinner wire in the ratio of 
55 to 43, or 1*3 to 1. Considering the minuteness of the quantities 
measured, and remembering that the thin and thick wires were 
probably not of the same material, we may regard these results as 
being in satisfactory agreement. 
The conclusion indicated so far seems to be that, at first, for 
small charges of hydrogen, the resistance of the wire increases pro- 
portionately to the charge ; but that, when the charge is great, the 
resistance is diminished. 
The experiments which form the subject of this paper were all 
made in the Physical Laboratory of the Imperial University, Tokyo, 
Japan. 
