1875,.! The Spectroscope in Mint Assaying. 89 
consideration, and with the view partly to search for unknown 
sources of error and partly to ascertain generally the quantity 
of metal operated on in a spectroscopic assay (should that 
ever be possible) the following experiment was tried :— 
Having weighed small electrodes, averaging eighteen milli- 
grammes each, with the greatest possible accuracy, on the 
gold assay balance of the Mint (which is sensitive to a 
twentieth of a milligramme, or even less), and having 
- arranged a spark register, it was found that Ioo0o sparks 
might be passed between these poles, each spark showing 
the spectrum of the metal distinctly, and yet the loss in 
weight was too small to be made the base of calculation. 
Thus, a gold pole lost in weight, after passing 1000 sparks, 
1-1000th of a grain; this gives for each spark I-1,000,o00th 
of a grain of gold, producing a bright spectrum. The 
number was then increased to 3000 sparks as a test. The 
loss of weight depends of course upon the electric volume, 
and in the experiments tabulated an endeavour was made 
to keep the latter constant. A slight deposit of the vapor- 
ised metal from the opposite pole takes place in fine 
division, but this is easily removed—in the case of copper 
and gold poles by dipping the gold for a moment in weak 
acid or by gentle rubbing. ‘The annexed tables, marked A 
and B, show that the loss in weight is marvellously small, 
averaging less than seven-tenths of a milligramme of gold 
for 3000 sparks. To give the amount for each spark, this 
must be divided by the number of sparks; thus, in round 
numbers, an electrode loses 1-1oooth of a grain after passing 
3000 sparks; or for to00 sparks 1-3000th of a grain, or for 
each spark I-1,000,o0o0th of a grain. The exceedingly small 
quantity of metal thus assayed renders this process, in the 
writer’s opinion, inapplicable to the operations in the Mint ; 
for it is necessary to determine gold assays to the 1-10,oooth 
part of the normal assay weight, and it is hardly conceivable 
that a discrimination to the 1-10,o00th part of the spark 
assay weight, or the I-10,000,000,00o0th of a grain, is prac- 
tically possible. Even if it were, it would not be proper to 
assume that a test on such an atomic scale would correctly 
represent the value of a large deposit, or even of gold ingots. 
It would certainly not be in the case of silver, which se- 
gregates. 
The experiments made by Cappel to determine the mnz- 
mum amount of each element that will show a spectrum 
have been published in tabular form. His method was to 
volatilise ‘‘ solutions of the metallic salts between the poles 
of a small induction coil in Mitscherlich’s glass tubes with 
VOL. V. (N,S.) N 
