14 



THE COMPRESSIBILITY OF LITHIUM, SODIUM, 



which follow. The change in the zero-point was of no consequence, pro- 

 vided that it did not occur in the course of a single series of experiments, 

 because a pressure of less than 30 atmospheres was never used ; the increase 

 in pressure was all that need be known. That the gauges were each trust- 

 worthy in any given series, many experiments demonstrated. The fact 

 that the value of compressibility of water as determined by the gauges at 

 the close of the experiments was found to be equal to that found at first 

 is, further, strong evidence of the constancy of the gauges. 



It is interesting to compare the value of the compressibility of a standard 

 substance, for example water, as computed with the help of these corrected 

 pressures, and the value for the same substances found by Amagat.* This 

 comparison is given in the table below, using an average of our three most 

 careful determinations, to be recorded later. 



The compressibility of zvater. 



From this table it is easily seen that if Amagat's standard was correct, 

 our 200 atmospheres space between 100 and 300 was less than one-third of 

 1 per cent in error, while the next 200 atmospheres were 2.5 per cent in 

 error. Amagat makes the total compression of water for the 400 atmos- 

 pheres 0.0166 of the volume, and we make it 0.01642. Whether these dif- 

 ferences were due to the possible defects in Amagat's apparatus already 

 mentioned, or to an incorrect standard used in graduating both our gauges, 

 it is impossible to say. 



Fortunately, this doubt, which would be a very serious matter in some 

 cases, has no effect upon the value of our result for the present purpose. 

 Relative compressibilities rather than absolute compressibilities are the 

 data sought ; and the gauges furnished relative results of an accuracy quite 

 sufficient. On this account it was deemed unwise to go to the very great 

 expense of time and money needed to procure a more accurate standard 

 of pressure. This is particularly true, because, as has been said, the results 

 below can very easily be corrected at any future time, whenever the differ- 

 ence between the compressibility of mercury and water is precisely kno\vn 

 with reference to a really accurate standard of pressure. 



*Ann. de Chim. et de Phys. (6), 22, 1891 (121). 



