60 THE COMPRESSIBILITIES OF THE MORE IMPORTANT 



which water served. Experiment 7, the compressibility of magnesium, 

 13.67 grams of metal hammered and turned down in a lathe to fit the 

 jacket were immersed in 0.843 gram of oil in jacket v. Enough mercury 

 \vas placed in the narrow limb of the jacket to cause contact to be made 

 under 103 metric pressure-units. Mercury was added in three successive 

 stages, until when 0.4149 gram of mercury had been added, contact was 

 made under 520 pressure-units. From the curve (or in this case, because 

 the pressure readings are near 100 and 500 respectively, directly from the 

 data) it is easy to see that 400 units of pressure correspond to 0.398 gram 

 of added mercury. Because the density of magnesium is 1.74, the calcu- 

 lation is as follows : 



/= (0.398-0.0^-0.4385X0.8.3)1.74 +0 _ 



or 



ft = 0.00000268 



Experiment i. The compressibility of zinc. 108.2 grams of pure 

 zinc was cast, hammered, and turned to fit jacket m, in which it was 

 immersed in 3.906 grains of water. The point of the first contact was 

 42 pressure-units. Successive amounts, of 0.2997, 0.6084, and 0.9139 

 gram of added mercury were found to correspond to 183, 339, and 499 

 metric pressure-units, respectively; and on taking out all but 7.5 mg. of 

 this mercury the pressure went back to 39 pressure-units, almost the same 

 as at first. The latter point is to be considered as more reliable than the 

 first, but the difference is too slight to have a great effect on the results. 

 From these figures it is easily found that the mercury needed between 100 

 and 500 pressure-units is 0.790 gram, and the calculation becomes 



or 



ft 0.0000015 



The figures given in the table on p. 61 were found or calculated in the 

 same way, so that further explanation is superfluous. All the data neces- 

 sary for the calculation are given in the table, except the values W cor- 

 responding to the various jackets. These data are to be found in the 

 table on page 58, entitled "Data concerning glass jackets filled with mer- 

 cury." The final results have each been recalculated at least three times, 

 in order to be sure that no error could have crept in during the somewhat 

 elaborate computation. 



Moreover, besides being calculated in the way described, they were also 

 calculated directly from the observed points in each case, reading the cor- 

 responding values on the water and mercury curves, without drawing 



