ME. G-EOBGrE GOEE ON ELUOBIDE OE SILVER.. 
243 
Pare chlorine was passed in a slow stream during 1-| hour over 30'66 grains of pre- 
viously fused argentic fluoride in a boat of purified gas-carbon in a short platinum tube 
within a long one at a barely visible red heat ; a fuming vapour, which corroded glass 
freely, was evolved during the earlier part of the process, and a small amount of platinum 
salt sublimed. The saline residue weighed 35 T4 grains, theory requiring 34’82 grains 
of argentic chloride, if the correction be made for the amount of free silver in the original 
fluoride. The residue was non-deliquescent, and had the physical properties of argentic 
chloride slightly coloured by traces of platinum salt. After cleaning the two tubes by 
means of aqueous ammonia and by hydrochloric acid, the short one had lost 3 ‘5 8 grains, 
and the long one l - 31 grain by corrosive action of the gas. In this experiment I con- 
sider the fluorine passed away in chemical union with the carbon of the boat, and that 
the corrosive action upon the platinum was due to the chlorine. 
69*59 grains of recently fused fluoride in a partly purified and recently ignited boat of 
Siberian graphite, was heated to incipient redness in the graduated platinum receiver 
apparatus during 3^ hours, and then stood all night. No leakage occurred. On opening 
the receiver under mercury less powerful rarefaction occurred than when a boat of pla- 
tinum or gold was employed, and about 260 cub. centims. ( = 12*1 grains of chlorine) of 
mercury ran in. A slow absorption of gas then took place, after which the residuary 
gas was 130 cub. centims. at ordinary pressure and temperature. On opening the receiver 
free chlorine was found, and the gas fumed in the air. A film of brown sublimed salt 
of platinum was spread over the inside of the vessel. The saline residue weighed 76 T 8 
grains ( = 14T1 grains or 302*27 cub. centims. of effective chlorine); it consisted of 
argentic chloride containing some undecomposed fluoride. After cleaning the platinum 
articles by suitable solvents, they were found to have lost 4*81 grains = 3*467 grains, or 
74*3 cub. centims. of chlorine taken up by them and rendered non-effective. Some 
chlorine was also absorbed by the mercury on its admission. The carbon boat was 
slightly corroded around the silver-salt, and its weight was increased 2*07 grains ; by 
heating it to low redness in a nearly closed platinum retort, it evolved a strong acid 
odour and a vapour which corroded glass rapidly, and lost 1*57 grain in weight. I 
consider that the residuary gas was a compound of fluorine and carbon containing free 
chlorine, and that there were interferences in the results caused by several circumstances. 
To obtain a more accurate result with the same apparatus, I took an excess, or 11711 
grains, of recently fused fluoride in a very pure boat of Borrodale graphite (weighing 
119 ‘4 grains) which had just been heated nearly to redness. A slight leakage took 
place, and the receiver had to be refilled with gas, thereby causing the silver-salt to 
take up *4 grain of chlorine. The end of the retort containing the boat was heated to 
low redness during four hours, and the apparatus then set aside thirty-six hours at 60° 
Fahr. On opening the receiver under mercury, 282 cub. centims. ran in under ordinary 
pressure^ 123 cub. centims. of residual gas. Warming the receiver did not much expand 
the enclosed gas, showing thereby the absence of any highly volatile liquid. On sepa- 
rating the retort from the receiver no chlorine remained, and the mercury was not 
