38 REPORT — 1875. 



pitate again with spiribs of wine, and add the small quantity of chloride of 

 platinum and potassium to the bulk." 



In reply to the above criticism the Glasgow chemists say : — "In his analy- 

 sis of this sample Dr. Fresenius followed the method described in the sixth 

 edition of his ' Quantitative Analysis ; ' but, evidently fearing that the digestion 

 of the precipitate with alcohol of even 80 per cent, might not free it from 

 sodium compounds, he used a little water, wherewith to ensure the separa- 

 tion of the latter, and afterwards estimated the potassium in the washings 

 obtained, adding this to the main result — a plan which of course can be 

 equally adopted with our process if considered necessary " *. 



Messrs. "Wallace, Tatlock, and Clark further write : — 



" Our method obviates the necessity of converting sulphates into chlorides 

 before applying the platinum process. All that is necessaiy in the case of 

 these salts, or where they are present, is to add an equivalent qiiantity or 

 rather more of pure sodium chloride, which takes up the liberated sulphuric 

 acid. We believe that the general tendency is to report potassium results too 

 high, not only on account of incomplete removal of sodium compounds from 

 the potassium precipitate, but by reason of impure platinic solutions, which, 

 however pure when originally made from the metal, are liable to contamina- 

 tion through the spent liquors and precipitates being recovered by question- 

 able means. 



" There is almost no limit to the accuracy of this process ; with care and in 

 good hands, the potash may be estimated easily to within -05 per cent." 



Mr. W. Galbraith, who has had great experience in the analysis of potash 

 salts by Mr. Tatlock's method, writes of it as follows : — 



"The method requires a few precautions, the principal of which are that 

 the pipette be accurate and that the platinic chloride be pure. Care also 

 should be taken that the evaporation should not go to dryness, especially in 

 presence of a large quantity of soda salts, or the result will be too high. By 

 diluting the solution previous to the evaporation, the precipitate comes down 

 in largex crystals, is more easUy filtered and taken off the filter, and is also 

 more likely to be pure. 



"With these precautions, which are easily attended to, the method gives 

 rigidly accurate results even in the hands of inexperienced manipulators." 



Dr. G. L. Ulex, of Hamburg, separates any sulphates by very cautious ad- 

 dition of chloride of barium. He washes the chloroplatinate of potassium 

 with alcohol of SO per cent. He obtains results reliable within -2 per cent. 

 The process, "although simple, requires to be worked carefully, otherwise 

 serious mistakes will be made." 



Mr. M. J. Lausdell says, "I consider the great secret is to use plenty of 

 platinum, which facihtates the washing out of the sodium salts, and renders 

 the indication by colour (as to when to arrest washing) distinct." 



M. Joulie separates sulphates, and then the barium introduced, together 

 with any calcium or magnesium present. He washes the chloroplatinate 

 with a mixture of alcohol and ether. Absolutely exact results are obtained 

 in .5 or 6 hours, whatever the nature of the original sample. 



Dr. G. Berrand uses a large excess of platinic chloride, and washes the 



. * As the quantity of platinic-chloride solution employed by Mr. Tatlock for washing 

 the preoipitat* does not exceed 70 fluid grains, the solution of any sensible quantity of the 

 latter seems improbable, and if occurring might be altogether prevented by previously satu- 

 rating the platinic chloride used for washing with the potassium salt. 



