C. M. Warren on Organic Elementary Analysis. 389 
shorter tube would answer equally well for many non-volatile 
bstances. It will be observed that the asbestus packing is but 
the bar nearest or most remote from the ame, Or an interme- 
diate point, according to the temperature required. 
he steadiness of the heat thus applied, and the facility with 
which it may be regulated by simply moving the bar, render it 
decidedly preferable to any other means which I have employed 
r that purpose. I had for a long time used such a bar for the 
Same purpose in the old process, with extreme satisfaction. In 
Some cases a bar of copper laid on the combustion furnace,’ one 
end projecting into the flame by which the tube is being heated, 
and the other end raised and extending toward the substance, 
has been found to answer a good purpose. 
2. In the case of volatile bodies (I have not yet analyzed any 
others by this process), I have found the combustion to 
most satisfactorily when, having first heated about four or five 
inches of the anterior portion of the tube, which includes the 
oxyd of copper, and started the flow of oxygen, I apply the 
ted bar to the bulb containing the substance, and immedi- 
ately expel the whole of the liquid,—which becomes at once 
* Tt has occurred to me that my safety-tube may serve as the basis of a more 
simple and equally accurate process for the analysis of gases by gradual com 
= instead of explosion, in which weighing would take the place of measurement. 
oO 
: fastened together with wire, no harm could ever occur from overheating. A expe 
Bohemian glass, thus protected, m: yses; 
indeed, become almost a ent fixture upon the aces ld a eh ooo 
Vents the glass and metal from adhering together,—which 1s proba! 
. : = fhe that sudden cooling and pte | 
It is important that the iron trough 
only cause of breakage of wrar 
May take place with perfect security. 
