of simple alimentary Substances, fyc. 35 



extraneous water be present ; but all pulverulent substances, 

 and oxide of copper among the rest, are more or less hygro- 

 metric, and rapidly attract moisture from the atmosphere. 

 This circumstance seems to have struck the French chemists, 

 and it occurred to me at a very early period. Dr. Ure, how- 

 ever, was the first who published a method of obviating this 

 difficulty ; and his method, if this were the only difficulty to 

 contend with, is capable of considerable precision. But there 

 is another, and perhaps still more troublesome property, pos- 

 sessed by the oxide of copper, in common with many other 

 powders, namely, that of condensing air as well as water * ; 

 and these two difficulties, taken together in conjunction with 

 another mentioned in a note below, render great precision 

 almost out of the question f. To conquer these, every means 



* See Saussure's paper on the absorption of the gases by different bo- 

 dies. Annals of Philosophy, vi. 241. Also Gilbert's Annalen der Phi/sick, 

 xlvii. 112. 



f As I am unwilling that so much labour should be lost, particularly as 

 it may be of some use to other inquirers, I have thrown into the form of a 

 note a few of the principal circumstances connected with the inquiry men- 

 tioned in the text. In my earlier experiments, tubes of iron, copper, &c. 

 were employed instead of glass, and charcoal instead of spirits, as the me- 

 dium of heat ; and during this period most of the modifications of apparatus 

 which have been since proposed as novelties or improvements, were tried 

 and rejected. I first took the hint of employing a spirit lamp from Mr. 

 Porrett, and was certainly among the first that did so employ it. Various 

 forms of lamps were tried, but at length I was induced to relinquish the 

 use of the horizontal apparatus for the vertical one a ; and this, I have no 

 hesitation in saying, is by far the best form of apparatus hitherto proposed 

 for the substances to which it is adapted ; nor would any other have been 

 required by me, at least, had it not been for the properties of the oxide of 

 copper alluded to in the text, which render this and all other forms of ap- 

 paratus depending on the employment alone of that substance perfectly 

 useless when great accuracy is required. It has been objected to the lamp, 

 that the heat produced by it is not sufficient ; but this is a mistake ,• at 

 least I have never met with any substance that resisted its action, provided 

 the oxide of copper was well shaken up in the tube, or, if necessary, taken 

 out of the tube and retriturated, and afterwards exposed to heat a second 

 time, one or other of which ought to be done in all instances, whatever be 

 the medium of heat employed j for no ordinary heat will induce the oxide 

 to part with its oxygen to a combustible substance at some distance off, and 

 not immediately in contact with itself. A great heat is also attended with 

 some disadvantages, and among others, that of causing the oxide to adhere 

 together in hard and solid masses, which thus beeomes more difficult to be 

 removed from the tube, and much less adapted for future experiments. In 

 general, organized bodies are more difficult of combustion, and require more 

 heat than crystallized ones. The lamp described in the text I have only 

 recently employed, and it answers the purpose in all respects better than 

 any I have yet seen. 



With respect to the sources of error above mentioned, it was found that 



* Described in the Annals of Philosophy, xv. 190 (O. S.) ; and more com- 

 pletely in Dr. Henry's Chemistry, ii. 167, ninth edition. 



F 2 that 



