16 DISPLACEMENT INTERFEROMETRY BY 



The next experiments were made after some improvements of apparatus 

 and carried out with the same rod through a range of about 15 kg. The rod 

 stood the stress well, except at the end, when it showed slow viscous shrink- 

 age. The data are given in figure 10, a and b, and in table i and contain the 

 running displacement A/V of the micrometer screw, as well as the successive 

 displacements Ae measured on the ocular micrometer. 



It has been found convenient for the sake of comparison with the preceding 

 set to compute E from three turns (5.46 kg.) of the screw, and A/V is so speci- 

 fied. For loads up to five additional turns (total 1 1 kg.) the data are practic- 

 ally identical, both in the outgoing and return series, and identical with the 

 preceding series. At six turns (total 13 kg.) the rod yields, but at seven turns 

 it again stiffens in both cases. As a whole the data are quite as good as the 

 reading of the micrometer screw admits. 



The values of &e are rough, for the fiducial mark of the screw was necessarily 

 taken near the middle of the field of the telescope, as a result of which the 

 large displacements Ae were thrown off to the edge of it, particularly in the 

 return series. Moreover, very small fringes were used for convenience, as 

 these were adequately mobile. Under these circumstances about A^=i3 

 ocular scale-parts correspond to io 3 AA^=i cm. To use Ae to advantage, a 

 special adjustment insuring the illumination of 100 scale-parts is desirable 

 and the fringes should be large. Only in case of more rigid rods where AW 

 fails is this necessary. 



7. The same. Continued. To determine the limits toward which E ulti- 

 mately approaches, it is advisable to examine a thinner rod; for beyond seven 

 or eight turns of the compressing screw, the stress put on the mechanism in 

 further twisting is liable to dislocate certain of its parts or to alter its position. 

 The convenience of the thumb-screw would be unavailable. A hard-rubber 

 rod of the same material was therefore made more slender, the dimensions 

 being, length , = 2.55 cm., diameter 0.215 cm -> ^ =0.0363 cm. 2 The ratio of 

 length to diameter thus exceeds 10/1. Large fringes were installed unneces- 

 sarily, for the sensitiveness of the interferometer is here excessive. The 

 dimensions given lead to 



if AJV is in lo- 3 cm. Omitting the earlier series, the results of the third and 

 fourth are given in table 2 and figure 1 1 , the twists of the compressing screw 

 being successively increased by one-third of a turn, or about 0.6 kg. There 

 was a small permanent initial load, as usual. 



The results as a whole are good. The ascending and descending branches 

 of the graphs are straight, except at the beginning or the end. The latter is 



