130 
Again, to the insulating piece (C) there is attached an arm (D) which 
supports one of the fibers of the bifilarly suspended mirror (B), the other 
fiber being attached to a fixed pillar (F), which has an adjustment to 
change the distance between the ends of the fibers. To explain the action 
of this mechanism, suppose the field magnet (M) to be excited, and thereby 
producing a very intense magnetic field in the field gap, and that a small 
current be flowing in the deflection coil freely suspended in this field gap. 
The result will be an upward or downward motion of the deflection coil, 
according to the relative directions of the currents in the two coils. 
This motion will be communicated by means of the terminals to the 
pulleys (P, P), which will be rotated about their axis, the piano wire, car- 
rying the arm (D) with them. 
It will be seen that if the ratio of the radius of the pulleys to the arm 
(D) be 1-r, then will the upper extremity of (D) describe an are whose 
length is r d, where d is the distance that the coil moves up or down. 
Again, suppose the horizontal distance between the upper ends of the 
bifilar suspension to be s, then will the mirror (E), which, of course, hangs 
in the plane of the two fibers, make the angle Tan-!rd-+swith its initial 
position. And therefore for a given value of rd. the deflection of the 
mirror will approach a maximum as s decreases. 
We are thus able to increase the sensibility of the instrument in a very 
convenient manner, and what is more, we can change it a given amount 
at will by simply changing the distance s by means of the graduated 
screw (T). 
The instrument has been completed but a short time, and therefore we 
have not been able to give it a fair test, but the experiments that have 
been made are sufficient to indicate that, with a few changes in the me- 
chanical details, our object has been attained. 
Note oN SOME EXPERIMENTS TO DETERMINE THE RATIO BETWEEN THE ELASTIC 
Limit iN TENSION AND IN FLEXURE FOR Sort STEEL. By W. K. Harr. 
The fact that the material at the top and bottom of a beam of ductile material 
will show an elastic limit in flexure higher than its elastic limit in tension has 
been noted by experimenters, by Baoshinger and M. Considere, for instance. For 
steel bars the elastic limit in flexure was 4 to } larger than in tension, and the 
increase is a function of the shape of cross-section and ductility of material. The 
