COPPER. 95 



And the percentages are: 



Water at 260°. At 370° . Cu in Cryst. Salt. Cu in CuSO^. 

 7 25.452 



s 35-970 25.446 39-740(260°) 



9 36.067 25.445 39.799(370°) 



25.44S 



In this series the determinations of sulphuric acid gave essential!}^ the 

 same results for all three samples of sulphate, although one was not 

 dehydrated, and the others were heated to 260° and 370° respectively. 

 Hence the loss of weight in dehydration at either temperature represents 

 water only, and does not involve partial decomposition of the sulphate. 

 Jjetween 360° and 400° copper sulphate is at'essentially constant weight, 

 Ijut further experiments indicated that even at 400° it retained traces of 

 water, and possihly as much as .042 per cent. The last trace is not ex- 

 polled until the salt itself begins to decompose. 



Richards also effected two syntheses of the sulphate directly from the 

 metal by dissolving the latter in nitric acid, then evaporating to dr3"ness 

 with sulphuric acid, and heating to constant weight at 400°. 



.67720 grin. Cu gave 1.7021 grm. CuSO^. 39-786 per cent. Cu. 



I. 00613 " 2.5292 " 39.781 



If we include these percentages in a series with the data from anal^'ses 

 4, 6, and 9, which gave percentages of 39.811, 39.799,and 39.799 respect- 

 ively of copper in sulphate dried at 360° and upwards, the mean l^ecomes 



CuSO^ : Cu : : 100 : 39.795, zfc .0036 



Since even this result is presumably too low, the other figures from 

 sulphate dried at 250° must be rejected. Since Hampe's work on the 

 sul{)hate is afiected by the same sources of error, and apparently to a 

 still greater extent, it need not be considered farther. As for Richards' 

 nine determinations of Cu in CuS0^.5H^0, we ma}^ take them as one 

 i«eries giving a mean })ercentage of 25.451, ± .0011. This salt seems to 

 retain occluded water, for the percentage of copi>er in it leads to a value 

 for the atomic weight which is inconsistent with the best evidence, as 

 will be seen later. 



In the second and third series of Richards' ex])eriments upon copper 

 sulphate, the sulphuric acid was estimated b}' a method which gave 

 valuable results. After the copper had been electrolytically precijDitated, 

 the acid which was set free was nearly neutralized by a weighed amount 

 of pure sodium carbonate, and the slight excess remaining was deter- 

 mined by titration. Thus the weight of sodium carbonate equivalent to 

 the copper was ascertained. The resulting solution of sodium sulphate 

 was then evaporated to dryness, and a new ratio, connecting that salt 

 with copper, was also determined. The cross ratio Na.,C03 : Na,,SO, has 



