1020 
PROFESSOR J. W. MALLET OR A REVISION 
hard glass, previously closed in at one end, carefully dried, and sealed at the other as 
soon as the tube was nearly filled, the fused bromide having been introduced through 
a miniature tube funnel to avoid smearing the upper end of the collecting tube, and a 
new, perfectly dry funnel used each time. The weight of each collecting tube had 
been taken beforehand, and the piece of glass drawn off in sealing being washed, 
dried, and weighed, the weight of the sealed tube itself was known, the difference 
between this and the total weight of tube and contents giving the amount of bromide. 
Experiments to determine the amount of bromide in this compound by precipitation 
with a silver solution have the advantage over those of Dumas, above quoted, upon 
the chloride that, as Stas has shown,* silver bromide practically does not share with 
the chloride of this metal the slight solubility in the exactly neutralised liquid from 
which precipitation has been effected which renders difficult an exact determination of 
the amount of silver needed. Pursuing in general the course so carefully examined 
by Stas, the following were the details of the method employed. 
Preparation of pure metallic silver .—Pure metallic silver was prepared by dissolving 
in nitric acid nearly pure silver already on hand, diluting largely, precipitating with pure 
hydrochloric acid, digesting the precipitate with aqua regia, washing thoroughly, and 
reducing the purified chloride in the liquid way with sodium hydrate (from metallic 
sodium) and invert-sugar (from perfectly pure and well crystallised cane-sugar boiled 
with dilute hydrochloric acid). The metal, after having been carefully tested, was 
fused by a jet of purified hydrogen mixed with rather less oxygen than necessary for 
perfect combustion. To avoid the necessity for any cutting up afterwards with steel 
tools, the pulverulent metal was divided into a number of little lots of various weight, 
and these were supported upon the surface of little blocks made by compressing pure 
sugar charcoal (from cane-sugar quite free from heavy metals) made into a paste with 
pure cane-sugar syrup and gradually drying and heating to redness.t The fused silver 
thus obtained was examined for occluded oxygen, following the method of DumasJ in 
his recent experiments. It was supported upon a thin layer of pure lime in a hard 
glass tube, and heated to moderate redness in a Spiif.ngel vacuum. The amount of 
oxygen given off was less than that obtained by Dumas, doubtless owing to the fact 
that his silver was fused under nitre and borax, while mine was, as just stated, melted 
on a surface of carbon with no flux. He obtained at the rate of 57 c.c. of oxygen 
(for 0° C. and 760 m.m.) per kilogramme of silver, and in other experiments, prolong¬ 
ing notably the time of fusion, as much as 158 c.c. and 174 c.c. I obtained but 
34'63 c.c. per kilogramme, and in another experiment made by Mr. Santos, then a 
student in this Laboratory, the silver having been fused upon ordinary wood charcoal, 
but 30T2 c.c. was given off All the silver used in the atomic weight determinations 
* ‘ Comptes Rendus,’ 73, 998. ‘ Annales de Chimie et de Physique ’ [5], 3, 289. 
t This form of support had the advantage that if any particles of carbon should be mechanically 
enclosed in the silver they would be readily seen on solution of the latter in nitric acid. 
J ‘Comptes Rendus,’ 86, 65. ‘ Chem. Centralblatt,’ 27 Febr., 1878, S. 138. 
