359 
of simple alimentary substances , &c. 
elaborate paper on the subject of vegetable analysis, in which 
he likewise employed the chlorate of potash, but in quite a dif- 
ferent manner ; and to this celebrated chemist I believe we are 
indebted for the improvement subsequently adopted by most 
of his successors, of introducing the mixture of the substance to 
be analysed, and the oxide, into a narrow tube, and submitting 
the different portions of it to heat in succession. The results 
of Berzelius were in general more accurate than those of his 
predecessors, especially as far as related to the quantity of 
carbon, but his method was not well adapted for determining 
the proportion of hydrogen.* In 1816, Gay Lussac seems 
to have thought of employing the oxide of copper for the 
purposes of analysis,-f the introduction of which undoubtedly 
constituted one of the greatest improvements hitherto made 
in organic analysis ; and the use of which has continued to 
the present time, and will perhaps never be entirely super- 
seded. The oxide of copper has however some disadvantages, 
which it is one object of the present remarks to point out; 
another is, to propose a form of apparatus free from most of 
the objections to which those hitherto in use have been more 
or less liable. 
There are two methods of arriving at the quantity of water 
formed during the combustion of an organized substance ; 
either actually to collect and weigh it, as Berzelius did, or to 
estimate the quantity by the loss of weight sustained by the 
tube after the combustion. The latter in general is the best 
method, and was that adopted by me from the first : it has 
since been followed by Dr. Ure, and others.J Whichever 
* Annals of Philosophy, iv. 323. f Annales de Chimie, xcvi. 306. 
J Phil. Trans. 1822. p. 457. 
