64 THE COMPRESSIBILITIES OF TPIE MORE IMPORTANT 



daily calcium and arsenic. Several others are somewhat less, although 

 more than half as large, namely, magnesium, zinc, cadmium, lead, thal- 

 lium, antimony, and bismuth. Tin, aluminum, sulphur, manganese, and 

 chromium are only about one-fifth as compressible as mercury, and copper 

 is still less compressible. Gold, palladium, and iron are yet less affected 

 by pressure ; and platinum, nickel, and molybdenum have a compressibility 

 not much more that one-twentieth of that of mercury, or one two-hun- 

 dredth of that of water. The determination of these very small quan- 

 tities is of course subject to considerable percentage of inaccuracy, but at 

 least it seems probable that the values given are not in excess of the true 

 values. 



It becomes now a matter of interest to compare the compressibilities of 

 the different elements in a more careful and systematic manner. This is 

 most conveniently done by arranging all the values which have been found 

 for the solid and liquid elements in a table in the order of the atomic 

 weights. Such a table is given below. In it are recorded not only the 

 compressibilities reported in the present paper, but also those found by 

 Richards and Stull, three years ago, and those concerning the alkali metals, 

 and certain non-metals given in the preceding papers. Of these deter- 

 minations two, namely, those concerning sulphur and selenium, were deter- 

 mined by us at the same time as the elements discussed above. They were 

 printed in a foregoing paper merely in order that all the work might be 

 arranged as systematically as possible. In this table of compressibilities 

 the values are given in terms of the megabar a megadyne per square 

 centimeter as the unit of pressure, instead of the kilogram per square 

 centimeter. This absolute unit is about 2 per cent greater than the latter- 

 named technical unit, and 1.3 per cent less than the "atmosphere." In 

 most of the present cases the difference between these standards does not 

 exceed greatly the experimental limit of error. The values may be easily 

 transposed into terms of the atmosphere by adding in each case 1.3 per 

 cent. In comparing these results it must be borne in mind that all depend 

 upon the still somewhat uncertain compressibility of mercury, and that 

 any change in this latter value would affect each other value by the same 

 actual amount (not the same percentage amount). No change of the 

 value for mercury could affect the sequence in the order of magnitude of 

 the results, which for the present purpose is the really essential matter. 

 There are given also the atomic weights, the atomic volumes of these 

 various substances. 



