FLAME SPECTRA AT HIGH TEMPERATERES. 
205 
Chromium. 
The spectrum obtained from ferro-chrome containing 22 per cent, of chromium, 
contains six lines due to chromium, and in addition bands and lines of iron. The 
bands extend from 24 to 28‘3, and continue weaker as far as 35. Manganese lines 
are also very strong. 
Ivory scale 
numbers. 
Description of 
spectrum. 
1 
\ • 
Remarks. 
O 
O 
00 GO 00 
A group of three 
well-defined lines 
r 2331 
2338 
[ 23-50 
4290 
4277 
4255 
Angsteoh and Thalex. 
4289-4 
4274-6 
4253 9 
1.32-2 ] 
133- 3 } 
134- 9 J 
A grou]3 of three 
well-defined lines 
r 27724 
27810 
1 27935 
3607 
3595 
3580 
Liveixg and Dewae. 
3606 
3593 
3578 
Chromic Trioxide. 
This substance gives, in addition to the above, two groups of three lines, a con¬ 
tinuous spectrum, strong, from close to the sodium line in the yellow, but a little less 
refrangible up to X 3820. 
Iridium. 
This element occasioned some difficulties. Strips of iridium, twisted into loops, 
were obtained from Messrs. Johnson and Matthey some years ago for the purpose 
of serving as supports for the alkalies and alkaline earths in the oxy-hydrogen blow¬ 
pipe. To this use it was put with some success and found convenient, but with 
oxides capable of undergoing reduction, even such as cupric oxide, it became corroded. 
It was found to be a convenient support for silicates which are fusible, but on 
examining the spectrum of silica, several lines were discovered which were not due 
to silica. 
Three varieties of silica were tested—1st, Silica precipitated from sodium silicate. 
This yielded lines identified with iron even after treatment with hydrochloric acid, 
2nd, Silica precipitated from silicon fluoride by passing the gas into water. The 
silica was evaporated from the hydrofluosilicic acid by filtration through absolutely 
pure ashless filter-paper. Even this showed a number of lines which at first were 
taken to be those of iron. 3rd, Rock crystal exposed to the hottest part of the 
flame on iridium for one hour gave nothing beyond the sodium lines in the yellow, 
mean X 5892, and in the ultra-violet X 3303. 
To prove the origin of the lines which had been previously observed, a piece of 
