PROFESSOR J. W. MALLET ON THE ATOAIIC WEIGHT OP GOLD. 
405 
weighings were made by observation of the oscillations of the index on either side of 
the position of rest. A difterence of weight of •0001 grm. with the smaller balance, 
and '0002 grin, with the larger instrument, was easily and distinctly observable with 
any load which the research required. 
The same kilogramme weight was made the basis of a comparison with all my other 
weights which had been before used in the same way. This had been compared at 
Washington with the “star kilogramme” of the United States Coast Survey, the 
value of which is known in terms of the original “ kilogramme of the Archives ” at 
Paris. All the smaller weights were carefully rechecked against this and against each 
other, and them real values ascertained as referred to a vacuum. The necessary 
determinations were made of the specific gravity of all materials and vessels which 
were to be weighed, and the barometer and thermometer were read at the time of 
each weighing, so that all weights recorded in this paper represent real values in vacuo. 
In order to reduce to a minimum errors due to varying deposition of hygroscopic 
moisture, vessels of like material, shape, and size with those used to contain substances 
to be weighed were used as tare. 
History and Mode of Purification of the Gold used in this Research. 
Most of the metal needed was prepared by myself, with precautions presently to be 
mentioned; a part was obtained, as “ proof gold,” from the United States Mint at 
Philadelphia ; another part from the United States Assay Office at New York ; and a 
single specimen of English “ trial plate ” gold from the Royal Mint in London. 
1. Purification of Gold hy the Author. —It may fairly be concluded from the general 
history of the gold of commerce that the impurities most to be stlspected, aiid most 
requiring special precautions for their removal, are silver and the metals of the 
platinum group. My joreliminary experiments led me to believe that the greatest 
difficulty in the way of obtaining perfectly pure gold consists in getting rid of the last 
traces of silver, the chloride of this metal not being quite insoluble in a solution of 
auric chloride. For the removal of silver, I have chiefly depended upon evaporation 
of the gold solution with a little hydrobromic acid, followed by large dilution with 
water, and long continued clearing by subsidence. As regards the platinum metals, 
my results agreed substantially with those of Hoffmann and Kruss,* but I have 
been inclined to lay some stress on reduction of the gold from its solution with exclu¬ 
sion of light, and on fractional reduction, using only the middle portion thrown down. 
I avoided altogether the use of ferrous salts as reducing agents, in view ol the difficulty 
of preparing them in large quantity with assurance of their purity, and the trouble of 
thoroughly washing the precipitated gold. For the final precipitation of the gold 
formic acid seemed to offer real advantages ; its volatility admits of easily getting it 
* ‘ Liebig’s Amialen,’ vol. 238, p. C6. 
