5 8 Mr. Smithson on the Composition of the 
Antimony - 25 . . . = 
Copper - !3 t • • • = To 
and it is not a little remarkable, that here, as was the case 
with the calamine,* they are sexagesimal fractions of it. 
When in a former paper I offered a system on the propor- 
tions of the elements of compounds, I supported it by the 
results of my own experiments, which might be supposed 
influenced, even unconsciously to myself, by a favourite hypo- 
thesis, and I made the application of it principally to a substance 
whose nature was not very clear. But the present case is not 
liable to these objections : here no fondness to the theory can 
be suspected of having led astray, nor did even the experi- 
ments as they came from their author's hands, bear an ap- 
pearance in the least favourable to it, and yet when properly 
considered, they are found to accord no less remarkably with 
its principles. 
It is evident that there must be a precise quantity in which 
the elements of compounds are united together in them, other- 
wise a matter, which was not a simple one, would be liable, 
in its several masses, to vary from itself, according as one or 
other of its ingredients chanced to predominate ; but chemical 
experiments are unavoidably attended with too many sources 
of fallacy for this precise quantity to be discovered by them ; 
it is therefore to theory that we must owe the knowledge of 
it. For this purpose an hypothesis must be made, and its 
justness tried by a strict comparison with facts. If they are 
found at variance, the assumed hypothesis must be relin- 
quished with candour as erroneous, but should it, on the con- 
trary prove, on a multitude of trials, invariably to accord with 
* Phil. Trans. 1803, p. 12. 
