500 
PHYSICS: C. BARUS 
Proc. N. a. S. 
indefinitely smaller and is subject to simpler equational conditions, but 
the apparatus itself may be made less cumbersome and steadier. Thus 
in figures 1 (elevation) and 2 (plan), r and r' are the two light rays of the 
interferometer. If w is the wire stretched between torsion heads (counter- 
poise W) and running normally to the balance beam, m, m', the auxiliary 
mirrors of the interferometer attached to the beam, the latter may be 
prolonged and terminate in the thin metallic plates, p, p'. These are 
surrounded by the narrow glass cases c', to secure air damping. This 
device is very efficient; and if c and c' are each placed on three leveling 
screws acting upward from below, while a stout vertical spring presses the 
cases downward from above (not shown), it is easy to free the plates 
p, p', particularly as they may now also slide, with slight friction, on the 
leveling screws. Hence, as the beam is practically aperiodic, on putting 
a wide brass plate underneath the system p m m' p' and close to it, no 
further casing is needed. The beam is thus easily accessible without 
opening the doors of a case. The fringes occasionally move, but they 
at once return to their equilibrium position, which is thus easily recognized. 
I tested this by removing the fixed counterpoise W, and proceeding with 
the method of weighing by exchanges. For this reason stiff hooks, k, 
figure 3, were fastened rigidly at each end of the beam, m. The plates p, 
p','mc, c', were soldered to the bottom of h as shown. The counterpoising 
weight, h, was bell shaped and reposed on the sharp point of the hook, 
h. A small projection on top allowed its easy removal with forceps. 
Adjusting the stops of the torsion head of the wire Wj the bell weight h 
could be passed (with the requisite torsion) from one end of the beam 
to the other, without any inconvenience, the fringes in either case of ad- 
justment appearing at once, in their proper position relative to the cross 
wire at the slit of the collimator. Of course, h must be stout enough to 
be free from flexure, relative to h. 
Observations. Twists Alternately Opposite. — If the wire is too thin, 
the fringes are liable to wander. This ceases entirely with a wire 0.05 
cm. thick, and about 35 cm. long, one-half of which is effective. It was 
of steel, hard drawn, but annealed to a blue temper to diminish the molecu- 
lar instability. A weight of the order of 1 gram was selected and the double 
twist produced by this was = 50°. The weight was passed from end 
to end of the beam, counterpoised by torsion, without difficulty and the 
fringes set to coincide with the shadow of the cross wire on the slit. Great 
care had to be taken to avoid twisting the frame carrying the wire and 
torsion heads. 
The constants of the interferometer being, angle of rays to mirrors, 
i = 45°, breadth of the ray parallogram, 6 = 10 cm., micrometer dis- 
placement AA^, we may write 5g/g = bB/B = AN cosi/b{AB/2) = 0.16 AN 
and as AN may be read to within 10 ~^ cm., the sensitiveness is about 
10 the gram counterpoise in question presupposed. The error to be 
