for the Analysis of Organic Bodies. 235 
shall just touch the tube, and only one set will be required, unless the 
tube be large, as for instance, half an inch in diameter * ; the lamps 
are to be lighted in succession, those nearest the gazometer first. 
If the substance to be analyzed be a vegetable salt, or be hy- 
grometric, it must be dried, which is best done in vacuo, but which 
Mr. Cooper effects also in the following manner. A wide-mouth 
stoppered bottle is selected, and alsoa smaller bottle which will 
easily go into it; a quantity of dry pulverized chloride of calcium is 
then strewed over the bottom of the larger bottle, and the smaller, 
containing the substance to be dried is also introduced; a small 
piece of bibulous paper is moistened with alcohol and put inside the 
larger phial; it is then lighted, and when it has burned for a second 
or two, the stopper is put into the bottle, and the vacuum obtained 
is such that the desiccation goes on very rapidly and effectually. 
When substances of this kind are analyzed, they must, of course, 
be-mixed with oxide of copper before they are introduced into the 
tube ; a quantity of pure oxide is then to be put into the tube, and it 
is as well to add afterwards a small quantity of copper filings 
orshavings. In heating the tube the wicks are to be lighted as be- 
fore, but instead of suffering the whole to burn at once, it is as well 
to leave only three or four in action at a time, extinguishing the 
others, but taking care to ignite the whole extent of tube at once at 
the end of the process. 
When nitrogen is present in the body to be analyzed it has a 
tendency to become oxidized at high temperatures by the oxide of 
copper, and in this case yields erroneous results. To obviate this 
as much as possible Mr. Cooper has lately used protoxide of cop- 
per, instead of peroxide ; and though he finds that in certain circum- 
stances this also will impart oxygen to the nitrogen, yet it does so 
with far greater difficulty than the peroxide: hence in all cases 
where nitrogen is concerned, the protoxide should be used. The 
protoxide is prepared by fusing peroxide of copper with copper 
filings in excess; a mass of protoxide is obtained, which, on being 
pulverized and sifted is fit for use, 
* The power of the lamps is such that a thick platinum tube, half an inch 
in diameter, may be rendered bright red-hot by them. 
