EFFECT OF TENSION ON CERTAIN ABNORMAL METALS. 45 



remained bright throughout the course of the experiment. The oil of 

 the bath, a neutral heavy white petroleum which is used for medicinal 

 purposes, was without perceptible chemical action, at least at room 

 temperature. The current and potential connections were main- 

 tained by mechanical contact, the potential connections by small 

 spring clamps of special design. 



The measurements of lithium were the most difficult of any, be- 

 cause of the extreme softness of the metal. Four sets of runs were 

 made in all. The first two, made without the thermostat, established 

 the sign of the effect and its probable magnitude. The last two runs, 

 with the thermostat at 30°, were somewhat more satisfactory. The 

 elastic limit is so low that I could not determine the linearity of the 

 effect within the elastic range. There was permanent stretch under 

 a load of 40 gm., and there were also irregular initial effects under the 

 first few grams of load, which may have been due to straightening of 

 the wire. All the best measurements were made between a load of 

 15 gm. as zero and 35 gm. The change of resistance under this maxi- 

 mum increase of load of 20 gm. is an increase of only 0.02%. The 

 extreme variation of the individual determinations for the best speci- 

 men was in the ratio from 7 to 20. Twelve determinations were made. 

 The probable error of the mean, calculated by least squares, was 6.7%. 



The tension coefficient of resistance, that is the proportional change 

 of resistance under a tension of 1 kg/cm^ was -i-4.9 X 10~^ for the 

 best specimen; the other run with the thermostat gave 4.6 X 10~^, 

 and the only one of the preliminary specimens which was worth com- 

 puting gave 4.7 X 10~^. In the following I shall assume for the most 

 probable coefficient +4.8 X 10~^. 



Young's modulus was determined from the bending of three speci- 

 mens, of the diameter given above and approximately 7 cm. long. 

 The maximum load applied to these specimens was 0.066 gm. Within 

 the limits of error the displacement was proportional to the load. 

 The values obtained for Young's modulus were respectively 4.72, 

 5.26, and 4.93 X lO^" Abs. C.G.S. units. Take as the most probable 

 mean 4.9 X lO^" C.G.S. or 5.0 X 10^ in kg/cm^. 



The cubic compressibility of lithium has been found by Richards ^ 

 to be 9.0 X 10~^^ pressure expressed in Abs. C.G.S. units. This may 

 be combined with Young's modulus by the formula of elasticity 



1 / Ek\ . 



a = -(l — ~^ I to find Poisson's ratio. The formula gives 0.42. The 



6 T. W. Richards, Jour. Amer. Chem. Soc. 37, 1643-1656, 1915. 



