OF THE ATOMIC WEIGHT OF ALUMINUM. 
1019 
By the following arrangement the bromide was prepared in large quantity and with¬ 
out any trouble. About 50 c.c. of bromine was placed in a large untubulated retort 
of hard Bohemian glass, the neck of the vessel standing vertically upwards, and an 
elongated piece of ingot aluminum, the upper end of which was firmly tied with 
aluminum wire to a glass rod, was cautiously dipped into the liquid and withdrawn 
as soon as violent action began. By alternately lowering and raising the glass rod 
the lower end of the metal was immersed in the bromine at intervals short enough 
to keep up the temperature of the latter and make the action practically continuous, 
while there was no actual ignition, and but little bromine vapour was lost. As soon 
as a considerable portion of this bromine had become converted into aluminum bromide 
the further action became manageable. The remainder of the main quantity of metal 
to be treated was now at once added in lumps of 10 to, 20 grms. each ; a long tube 
funnel with a glass stop-cock near the upper end was introduced into the neck of the 
retort, and liquid bromine was allowed to drip in at just such a rate as to keep up 
steady but not inconveniently violent action, taking care to keep the metal always 
covered. When the pieces of metal had nearly disappeared the supply of bromine 
was stopped, about 30 grms. more of aluminum was added in filings, the contents of 
the retort were digested for 4 hours at about 230° C., and the fluid portion was then 
decanted off from the insoluble residue into another (tubulated) retort. Most of the 
silicon was left undissolvecl as a brown amorphous powder. Most of the iron was 
converted into ferric bromide, which, during the continued heating, was in part broken 
up, leaving ferrous bromide instead. A little copper derived from one sample only of 
the aluminum used, was of course converted into bromide also. 
More than a kilogramme of crude aluminum bromide being thus prepared, it was 
purified by repeated fractional distillations at carefully regulated temperature, using 
as a receiver in each case the retort to be next employed, and adding each time, except 
the last, a few grammes of aluminum filings. About a sixth of the whole amount was 
each time first distilled off and rejected as liable to contain silicon bromide, a little of 
this compound actually occurring in the earlier distillations; and another sixth was 
left behind, in order to retain the iron, which was separated with greater difficulty. 
After five distillations the bromide was obtained perfectly colourless, and boiling 
steadily at 263°'3 C. under 747 m.m. pressure. Specimens were dissolved in water, 
and carefully examined for iron, silicon, copper, and other conceivable impurities, but 
none could be found. As an additional precaution, the last distillation was effected in 
a slow stream of pure nitrogen, so as to avoid any formation of oxide or oxy-bromide 
of aluminum, the propriety of this being suggested by Berthelot’s recent results* as 
to the thermic relations of aluminum to oxygen and the haloids, and the distillate was 
collected in three successive portions, the results of whose analysis will be separately 
given further on ; they go to show that these three portions were sensibly identical. 
The individual specimens of pure bromide required were collected in little tubes of thin, 
* ‘ Bulletin tie la Societe Chimique cle Paris,’ 20 Mars, 1879, p. 263. 
6 p 2 
