180 
MR. C. V. BOYS ON THE RYDIO-MICROMETER. 
Now, assuming that the working field is four times that found, then the product hi 
must be made one-seventh of that stated above, and the product E^. X H", which 
FiY-. 0. 
represents the available sensibility, must be made If times as great. It is satisfactory 
to find that the greatest efficacy, and therefore sensibility, is with the narrow circuit 
about six times as o-reat as that found for the wide circuit first considered. 
There is no reason why the exact dimensions assigned to the active bars should be 
employed; it will be well, therefore, to consider what will be the effect of using bars 
of other dimensions. 
Let the sectional area be supposed increased in the ratio 1 :n‘, then K will become 
wK, and C will become C/n; therefore, the greatest efficacy which depends on the 
product of these will be unchanged, but this assumes a constant difference of tempera¬ 
ture between the ends of the bars. Now, in the case of the increased sectional area, 
the radiation of heat upon the warm junction will be unchanged, while for a given 
temperature difference the flow of heat to the cool junction, both on account of 
ordinary heat conduction and the Peltier action of the current, will be increased, and, 
thus, the warm junction will not become so warm ; thus, the actual sensibility will be 
less. The loss of beat by radiation from the bars can only be less in consequence of 
the warm junction being cooler, and, thus, the conclusion remains true, that there is no 
limit, except that itnposed by the difficulty of working the materials, to the smallness 
of the sectional area that should be used. I may say here that I have found no great 
difficulty in making the bars as little as mm. thick, and in making perfect soldered 
joints where the weight of solder used does not exceed the fifth part of a milligram. 
