OF METALS OF THE PLATINUM GROUP. 
643 
This result proves that the alloy is not subject to liquation, and fully justifies the 
high opinion that H. le Chatellier and Roberts-Austen have formed as to its 
suitability for thermometric measurements. 
The process adopted for the separation and determination of the rhodium and 
platinum in this alloy is one that I have found by experience to be exceedingly 
accurate if conducted with care. (It is described by Turner, * Elements of 
Chemistry,’ 5th Ed., 1834, p. 652.) 
Fifty grains of each determination were dissolved in nitrohydrochloric acid, and 
evaporated to dryness. The dry chlorides were then dissolved up in a warm solution 
of sodium chloride, and again evaporated to dryness. The resulting dry mass was 
then taken from the evaporating dishes and triturated in a porcelain mortar, from 
which each determination was then transferred to closed flasks, and digested in 
absolute alcohol in a very slightly warmed atmosphere for 24 hours, the bottles being 
shaken from time to time. 
At the end of this period the whole of the platinum chloride is taken up in 
solution by the alcohol, and the rhodium chloride left combined with the sodium 
chloride as a double salt of rhodium and sodium chloride insoluble in alcohol. This 
latter salt was washed with alcohol to remove all platinum chloride, and then 
dissolved in water. 
From this solution the metallic rhodium was obtained by direct precipitation with 
pure zinc, washed digested in weak hydrochloric acid, collected, dried, ignited, and 
weighed. 
The platinum from the alcoholic chlorides when evaporated and re-dissolved in 
water was determined by precipitation with ammonium chloride and pure zinc, as in 
the palladium-platinum series, and weighed as metallic platinum. 
With these analyses accurately weighed portions of pure rhodium and pure 
platinum, in the proportions of platinum 900, and rhodium 100 parts, were dissolved 
and treated as standards to ensure accuracy in the results of the analyses. 
It will be observed that, in some cases of members of the platinum group, the 
results of the analyses add up in excess of 1000. 
This is to be explained by the exceptional difficulty which attends the accurate 
determination of the metals of this group. 
4 n 2 
