CRITICAL EXPERIMENTS FOR DETERMINATION OF POTASSIUM, ETC. 627 



intensified in the rubidium salt; and in its case exhibits itself strongly even 

 in the presence of a large excess of rubidium chloride. 



Our last Experiment, VIII., may perhaps be referred to as showing that 

 this irregularity can be avoided by using chloroplatinate of sodium as a precipitant, 

 and (may we add V) avoiding an excess of rubidium salt. 



When we were engaged in these experiments we had not yet discovered the 

 method of recrystallisation for the removal of surplus platinum, but an 

 experiment made incidentally with one of our chloroplatinates, viz., the one 

 whose analysis is quoted on page 621), proves that this chloroplatinate 

 did contain surplus platinum. 2 grammes of it were boiled with 100 c.c. 

 of water, the residue (a) filtered off hot, the filtrate, which was strongly 

 acid, cooled down, the precipitate formed by the cooling process filtered 

 off, and the filtrate evaporated to dryness on a water-bath. The residue was 

 treated with 8 c.c. of water, and the solution filtered. It gave no precipitate 

 on addition of chloride of platinum, but a strong precipitate on addition of 

 chloride of rubidium solution. The first residue (a) when boiled with water, 

 furnished a solution, which when evaporated, and (the residue) treated with 

 8 c.c. of water, gave a filtrate which produced slight precipitates with both 



reagents. 



V. Experiments on Chloroplatinate of Ammonium. 



For the preparation of the necessary supply of pure sal-ammoniac, we 

 started from a kind of commercial liquor ammonice, of 0'88 sp. gr., which had 

 been sold to us as having been prepared from volcanic sal-ammoniac. 300 c.c. 

 of this liquor were mixed with 65 c.c. of specially prepared hydrochloric acid 

 of 20 per cent., so as to neutralise about T \yth of the volatile alkali, and the 

 mixture heated, to drive out the still uncombined ammonia, which was 

 passed into 360 c.c. of the hydrochloric acid. The resulting liquid was alkaline. 

 It was evaporated in a Berlin basin over a water-bath, until a considerable 

 quantity of sal-ammoniac had crystallised out, the residue allowed to cool, the 

 magma of crystals collected on a funnel connected with a Bunsen pump, and 

 washed with small instalments of 10 per cent, ammonia solution, made from 

 the volcanic liquor, by distilling off ammonia from it, and passing it into water. 

 The salt was then transferred to a shallow basin, and kept under a bell-jar in 

 an atmosphere of (dilute) ammonia, over caustic soda sticks, which latter were 

 renewed from time to time. After about a fortnight, the sal-ammoniac was 

 transferred to another bell-jar, and under it, kept over oil of vitriol for 24 hours, 

 to remove the ammonia adhering to the crystals. After this last operation, the 

 sal-ammoniac was assumed to be pure, and was bottled up for the experiments. 



To make sure of its purity, however, a known weight of the salt (about 1*6 

 grms.), was dissolved in water to a known weight, and the chlorine determined in 



