MJR. C. V. BOYS ON THE RADIO-MICROMETER. 
163 
fibre, that, if any of them are made more or less, the value of the instrument will be 
diminished. As I have now completed the calculations up to a point beyond which it 
would be difficult to go, and where I believe but little on which the perfection of the 
instrument practically depends remains to be found, I wish, without further delay, to 
explain the formulae I have obtained. 
Owing to the large number of variables, and the complicated manner in which they 
are involved, it would be difficult by any direct mathematical process to find the best 
value for every one at the same time; what I have done is to take the variations, one 
or two at a time, in such an order that those taken later shall require little or no 
modification of the results previously found. 
Eig. 4. 
It is evident that the quickness of the instrument, whatever proportions may be 
given to it, will increase as the sensitive plate is diminished in thickness ; and later on 
it will be shown that, besides the quickness, the ultimate sensibility, he., the deviation 
for a given rate of radiation, may also be increased as the thickness is diminished. It 
may, therefore, be taken as a fact that, tlie thinner the plate, the more sensitive will 
be the instrument. When I began the calculations I did not think it likely that a 
plate composed of metals so difficult to \vork could be made much less than ^ mm. 
thick, and, accordingly, that has been assumed as one starting point. I have found, 
however, not the slightest difficulty in producing plates thinner than this, but I have 
in what follows taken this quantity as the tliickness when wishing to find the nume¬ 
rical values given by the formidm. 
The next thing to determine upon is the general size of the circuit. If it is made 
Y 2 
