PHYSICS: C. BARUS 
695 
supposing the bar F to rotate over an angle a, in consequence of the 
increment A P of thrust and A to be the corresponding displacement 
of the micrometer mirror (rays incident at the angle i), needed to re- 
store the interference fringes to their original position. But, 
M = R' ^a = R' ^N cosi / 2 R 
so that after reduction 
_ 2 Z i<: AP 
^ ^ A R' cos i AN 
The method will not of course be very precise, because for rods less 
than an inch long the quantities involved, particularly A N, are so small. 
Any flexure or slight dislocation of the parts of the apparatus are of 
relatively great consequence. Moreover there is another serious con- 
sideration. In long rods the stresses distribute themselves equally 
throughout the sectional area; but in short rods this is liable not to be 
the case. There will be lines of longitudinal stress ^nd part of the area, 
A, may be relatively unstressed. Hence the value of E will come out 
too small and the question is rather to what degree such a method can 
be made trustworthy. If the achromatic fringes are used, the optical 
method as such presents no difficulties. For reasonably thin rods the 
observed displacement is adequate. The fringes need not be counted 
and it is even unnecessary to make the method very sensitive. Fringes 
of moderate size suffice. 
The method of flexure would in some respects seem to be preferable. 
But it is is theoretically less simple and for short rods difficulties similar 
to the above would be encountered. 
2. Observations. — In a large number of measurements made with 
different bodies, the apparatus finally took the form shown in figure 1, 
in which the rod r, /, to be tested is held by a rigid metallic tube or 
sheath, in which it fits loosely. Even this can not be employed quite 
without misgivings; but these must be passed over here. As an example 
of the results, I will insert graphically the behavior of hard rubber 
rods, each of about 2.4 cm. long and 0.35 cm. in diameter, thus having a 
sectional area of about 0.1 sq. cm. and kept under a minimum load of 
1 kgm. These were subjected to cyclically varying stress with the 
results (contractions positive) given in figure 2. With pressures vary- 
ing in the sequence (1, 2, 3, 2, 1), (1, 2, 3, 4, 3, 2, 1), etc., kgm., the con- 
tractions apparently give evidence of well developed hysteresis loops, 
upon which is super-imposed the continuous viscous deformation which 
