328 
ME. GEORGE GORE ON FLUORIDE OF SILVER. 
experiment, and the excess of bromine then expelled by heat ; no sublimate occurred. 
The boat and its contents had increased in weight 22*39 grains, theory requiring 22*87 
grains if all the salt was converted into argentic bromide and its fluorine expelled. 
The platinum tube was not corroded, nor altered in weight. The boat had lost 2*52 
grains, apparently partly by corrosion; say, 2*25 by corrosion and -27 by moisture 
(as found in a similar experiment). The saline residue was translucent, somewhat 
malleable, insoluble in water, and weighed 71*58 grains, theory requiring 70*49 grains 
of argentic bromide, the slight excess of weight being accounted for by the fused 
fluoride having contained a little reduced silver. By fusing 5*85 grains of it with an 
excess of alkaline carbonates, 3*36 grains of silver, entirely soluble in dilute nitric acid, 
was obtained, theory requiring exactly that amount. I consider that in this experiment 
the whole of the argentic fluoride was converted into bromide, the fluorine escaping in 
chemical union with the carbon of the boat as in similar experiments with chlorine (see 
Phil. Trans. Boy. Soc. 1870, pages 243, 244). The bromine used in this experiment was 
not quite pure. 
I repeated the experiment with perfectly pure and anhydrous bromine, and a boat of 
Spanish graphite which had been perfectly purified by the process already referred to 
(page 327), the boat being heated nearly to redness immediately before the experiment. 
The boat after heating weighed 88*48 grains, and the fused fluoride 77*63 grains. The 
experiment lasted six hours, and much vapour, which corroded glass in damp air, 
was evolved. After expelling the excess of bromine, the boat was found slightly cor- 
roded, and, with its contents, weighed 201*82 grains. The saline residue adhered to the 
boat and could not be separately weighed ; its weight, however, must have been about 115 
grains, on account of the loss of weight of the boat ; it contained scarcely perceptible 
traces of soluble silver-salt. The results of this experiment perfectly agree with those 
of the previous one. 
It may be here remarked that gaseous fluoride of carbon does not corrode dry glass, 
and that the corrosive action of vaporous fluorine compounds upon glass is variable, 
and is caused in some cases by the compounds assuming the liquid state, and in others 
by the presence of traces of moisture. 
On adding liquid bromine to a saturated aqueous solution of argentic fluoride, an 
abundant precipitate was produced, but no gas was evolved ; on further adding fragments 
of the fluoride and stirring, effervescence occurred rather freely, and much heat and acid 
odour and a little oxygen were evolved. 
The reaction may be represented by the following equations : — 
8Ag F + 8Br + 4TI 2 O = 5Ag Br + 3Ag BrO + SHF + O, 
or 
7 A g Br + Ag Br 0 3 + 8HF + O. 
I now perfectly dried the yellow precipitate in a nearly covered platinum cup, and 
heated a portion of it with sulphuric acid in a glass test-tube ; bromine was liberated 
freely, but the glass was not corroded. With dilute hydrochloric acid the yellow powder 
