97 



the stream ignited, it gives a different flame, scarcely luminous, 

 but hotter than a candle urged by the blow-pipe, for I have fused 

 by it platina wire of tt- of an inch, and it may be remarked, by 

 this means we can ignite small quantities of precipitates, &c. to 

 whiteness in the most elegant manner ou a leaf of platina. Havino- 

 made the flame of whatever size we judge necessary, we gradually 

 admit oxygen, it changes its appearance, and is impelled with 

 much greater rapidity, diminishing in thickness, and becoming 

 more pointed and dense, while it emits a moaning sound. As 

 the supply of oxygen increases it becomes blue, surrounded by 

 a yellow cone, as in the connnon blow-pipe ; with more oxygen 

 the yellow cone disappears, and the flame becomes of a brighter 

 blue, it is reduced to a mere thread, and from the deficiency of 

 hydrogen, the brass of the jet burns while the heat is very much 

 diminished. It is not easy to ascertain the proportions of gas 

 which give the greatest heat, nor is it of much use, as a quantity 

 at its maximum is not much affected by a considerable change 

 in the variable on which it depends ; but I tried one experiment, to 

 know if it agreed with the theoretic proportion of 2 to 1. When the 

 heat seemed most intense, I extinguished the flame by turning the 

 hydrogen cock, and immediately returned it to its position, I then 

 collected the mixture which issued from the blow-pipe, 1 . 5 cubic 

 inches of it were exploded by the electric spark, and after the 

 requisite correction, it was found that there remained 0. 3, there- 

 fore 0. 4 of oxygen were present ; but my oxygen was impure, 

 containing i of nitrogen very nearly, therefore the residue must 

 have contained 0. 1 of this gas, and if the hydrogen was pure, 

 the result gives 10 hydrogen + 4 oxygen + 1 nitrogen, as near 

 as could be expected in this rough mode of estimation to the 

 VOL. xm. p 



