170 
MR. C. V. BOYS ON THE RADIO-MICROMETER. 
evidently is that, if the axial spike be removed altogether, as shown in fig, 5, the 
actual resistance and moment of inertia would not be so much as 1 per cent, different 
from the value (^) already found, and that, as these calculations are merely a guide to 
direct the design of a practical instrument, an assumption which is not more than I or 
2 per cent, in error in what is after all a small fraction of the total resistance and 
moment of inertia—one which, in fact, is so true that neither the power of the 
instrument maker nor our knowledge of the constants will prevent our introducing in 
practice variations of far greater magnitude—may be allowed to pass, especially as 
without something ecpaivalent it would be impossible to find expressions which would 
apply to the more complicated conditions which will be considered later. 
Fig. 5. 
The same considerations apply exactly to the sensitive plate, which, when the 
greatest efficacy with re.spect to moment of inertia is required, should no longer be 
made prismatic, but should become more narrow towards the ends, as shown in fig. 1, 
I may mention that in the original instrument, which I made before T had arrived at 
any of these results, I did, as a matter of fact, make the active bar of the shape 
sliown, because I felt that the lower corners were doing more harm by their moment 
of inertia than they were doing good by their conductivity. 
Since the active plate is taken as one of the invariables of the circuit, any propor¬ 
tions that may be desired may be taken for it, and tlie resulting resistance and 
moment of inertia made use of in calculating numerical values. I have in the 
arithmetical work which follows taken the active plate to be of tlie original shape and 
size, that is, a rectangular prism, 10 X 5 X ’25 mm. I have further not made any 
allowance for excess of metal for soldering, i.e., j) has been taken as the coeflicient of 
resistance and moment of inertia of the upper bar only, which has been found to be 
