OF ARTS AND SCIENCES. 261 



phuric acid and careful expulsion of the excess of this last hy heat 

 I have found it more convenient to mix the oxalate Ultimately with an 

 excess of amnionic sulphate and heat carefully in a muffle, as in the 

 last case. The resulting sulphates are perfectly soluble without pink- 

 ing together if sifted upon the surface of cold water. They are also 

 perfectly neutral. The oxalates may also be converted into chlorides 

 by mixing them intimately with amnionic chloride and igniting the 

 mixture in a muffle very gently. 



Determinations of mean atomic mass are now always employed in the 

 study of the mixtures of oxides which present themselves in the attempt 

 to effect separations. Very accurate results are obtained by the usual 

 method of converting a weighed quantity of oxides into the equivalent 

 weight of sulphates by treatment with sulphuric acid and subsequent 

 careful ignition. Probably this could be done more conveniently, and 

 in less time, by igniting the oxides in porcelain crucibles in a muffle, 

 after mixing carefully with amnionic sulphate, but upon this point 

 I have made no quantitative experiments. 



In all my work I have employed the analysis of the oxalates 

 convenient and accurate. It is, however, necessary to insist upon 

 several points of detail. In the first place I remark that the prepara- 

 tion of perfectly homogeneous mixtures of the oxalates requires much 

 care. It is best to precipitate by a hot dilute solution of oxalic acid 

 added slowly in small but distinct excess to a hot dilute solution of 

 the mixed chlorides or nitrates. The precipitated oxalates are then to 

 be thoroughly washed by decantation with hot distilled water. This 

 requires in general a large quantity of water, and must be continued 

 unLil the washings contain no trace of oxalic acid and give no cloudiness 

 with ammonia. Auer von Welsbach's process, which consists in adding 

 a very dilute solution of the nitrates (or chlorides) to a hot dilute 

 solution of oxalic acid, gives the oxalates in a state of \cv\ line >ul>- 

 division and perfectly free from hard masses. The mixed washings 

 on saturation with ammonia sometimes give a precipitate of oxalate s, 

 though in small quantity. These oxalates may be washed and mixed 

 with the main portion. The mass is to be dried upon a water-bath, 

 and then thoroughly mixed in a dry mortar. Only in this manner 

 is it possible to obtain a mass of oxalates sufficiently bomogeni 

 to yield corresponding results when differenl portions are analyzed. 

 The determination of the mean atomic mass in the oxalates prepared 

 as above involves only the determination of the percentag ■> oxide 

 R 2 0, and of C ? O g , the water present being "f course without influence. 

 Here I may remark that, as has doubtless been observed by other 



