PROFESSOR J. W. MALLET ON THE ATOMIC WEIGHT OF GOLD. 
415 
dried, first over sulphuric acid and afterwards over phosphorus pentoxide, until it 
ceased to lose weight; towards the end of the drying the temperature of the vessel 
was raised to about 50° 0. Preliminary experiments seemed to indicate the probable 
existence of this salt crystallized with a single molecule of wmter, but most of that 
prepared contained no constituent water, and it appeared easy to attain complete 
drying without any decomposition of the salt itself. Throughout its treatment the 
salt, which was not in any high degree hygroscopic, wms well guarded from dust and 
from any possible decomposing effect of light. 
The portion of the salt to be used in eacli experiment was contained in a small 
glass-stoppered weighing flask, which was removed just before it wms needed from the 
phosphorus pentoxide desiccator, the stopper having been inserted ; the flask was 
weighed, the greater part of its contents transferred quickly to a weighed porcelain 
crucible, the stopper at once replaced, and, the flask being again weighed, the 
quantity of gold salt taken from it was found by difference. 
In order to avoid mechanical loss by spattering on igniting the crucible and its 
contents, the auri-chloride lying together at the bottom of the crucible was covered 
by a layer, nearly a centimetre deep, of clean, carefully purified, and just previously 
well-ignited siliceous sand, the weight of this sand being known by taking it from a 
weighing flask in which it had been cooled over phosphorus pentoxide, and noting 
the loss of weight of this flask. In applying heat to the crucible and its contents it 
was found necessary to heat gently for a long time, raising the temperature slowdy, in 
order to prevent extensive charring at the bottom. Then, before the temperature 
had become too high, but after a considerable part of the volatile matter had been 
driven off, the sand was carefully stirred in with the remaining material so as to 
produce pretty uniform mixture, in order that the gold might not undergo partial 
welding together at a higher temperature, which might have led to wrapping up 
particles of carbon and their protection from combustion. In this operation a very 
small porcelain stirrer was used, as a platinum wire would have welded on and taken 
up some of the metallic gold ; the weight of this stirrer was determined in advance, 
and checked after use. Finally, the contents of the crucible were submitted to very 
careful and prolonged heating to moderate redness, with free access of air and occa¬ 
sional cautious stirring, so as to burn away every trace of carbon. After cooling in a 
desiccator, the crucible and its remaining contents were weighed, giving the weight of 
the i-esidual gold by subtraction of the weights of the crucible itself and the siliceous 
sand. As an additional safeguard against any particles of carbon left unburned 
escaping detection, the gold was afterwards dissolved out with aqua regia, and the 
white sand carefully looked over with a lens. 
The results of five experiments thus conducted were as follows :—■ 
