18 DISPLACEMENT INTERFEROMETRY BY 



0.205 cm - at one en d an d - 2I 5 at the other; part of a hard-brass nail. The 

 section was therefore A = 0.035 cm 2 . Hence equation (3) becomes 



E = nX 1.78X10^ -- 



0.35 



If AN is given in io~ 3 cm., 

 (5) E=io u X 



If the ocular micrometer shows a displacement Ae (in case of brass the 

 micrometer screw is insufficiently sensitive) equation (5) is to be modified by 

 first finding Ae/A/V by direct comparison. 



The tests gave an average value of Ae/AAfX io 3 = 17.1. Consequently 



ii7- n n 



io- u E =1.52X17. i =26 



A0 Ae 



for n turns of the compressing-screw. 



The behavior of the brass rod was peculiar and the first two series for 

 increasing and decreasing pressures are shown in figures 12 and 13. The 

 mean data are 



per turn 9.6 5.0 



io~ n 2.7 5.2 



5-3 5-2 



4-9 5-o 



Series (i) F increasing (i) F decreasing | (2) F increasing (2) F decreasing 



The first compression seems merely to have established firmness at the 

 seats at the end of the rod. Thereafter the value of E comes out fairly uni- 

 formly, but it is only one-half the normal value; i.e., &e is twice the reasonable 

 value, so that something else is simultaneously yielding. On the removal of 

 pressure there is at first no apparent elongation of the rod, but rather a con- 

 traction. The maximum load was about 16 kg., beginning with 2 kg. There 

 is no evidence of viscosity, but rather of permanent contraction or set. The 

 apparatus was vigorously tapped before each reading, which was thereafter 

 constant. The optical system functioned admirably. A piece of thin mica 

 placed under the contact lever showed A.N = 0.0229 cm. and therefore A# = 

 0.01786 cm. The same piece calipered with a micrometer screw proved to be 

 0.0180 cm. thick, practically a coincident result. 



The next four series were made with a higher initial load (4 kg.), but showed 

 the same character. In No. 3 figure 14, there was an accidental dislocation 

 owing to the endeavor to use a wrench at the thumb-screw. If we take mean 

 results for the outgoing and return series the data are 



Series 3 4 5 6 



Ae per turn 4.7 4.4 4.7 4.0 



io- u 5.5 5.9 5.5 6.5 



The normal value of the modulus is being slowly approached. As the dis- 

 placement is too large, errors can not be referred to friction. Probably the 

 two ends of the rod are fitting themselves to a more uniform seat. 



In the next series (8) figure 14, the projecting head of the rod was ground 



