346 Harriry AND RAMAGE 
Banded Flame-Spectra of Metals. 
As in the zinc-spectrum, the bands can be distinguished as far as the edge of 
the plate (A 3530), but the component lines of the bands are too nebulous and 
feeble to admit of accurate measurements. The lines are much more widely 
separated towards wave-length 3530. The lines in the spark-spectra are not 
identical with those in the flame-spectrum, though several of them have nearly 
the same wave-length. 
Fiame-Spectra oF ALUMINIUM, GALLIuM, INDIUM, AND THALLIUM. 
The oxyhydrogen flame-spectrum of aluminium, really of the metal burning 
in the flame, is a channelled spectrum of fine flutings degraded towards the red* 
It contains the bands which occur in the “ are-spectrum of aluminium oxide,’’t 
and in the spark-spectrum under certain conditions. t 
When aluminium burns in the oxyhydrogen flame, the energy of the 
combustion is sufficient to volatilise some of the metal, and the flame is coloured 
blue. This colour is also seen in the flame when oxide of aluminium, in the form 
of thin rods, is heated in the hottest part of the flame. A stronger spectrum 
was obtained when a mixture of aluminium oxide, with a dense form of carbon, 
such as gas carbon or graphite, was heated. The doublet of aluminium, wave- 
lengths 3967 and 3946, was photographed by this method, but the bands have not 
been obtained in the same way. 
Hemsalech confirms Aron’s view, that the bands are really due to the metal, 
and not the oxide of aluminium.$ 
No bands have been observed by us in the spectrum of gallium,|| but the metal 
is so scarce, that no opportunity has occurred in which to examine the spectrum 
thoroughly. 
Flutings not previously observed have been photographed in the flame- 
spectrum of indium, and in that of thallium; they are very much weaker than 
the lines of the strong doublets of indium with wave-lengths 4511 and 4102, 
and of thallium 5314 and 3775. 
* Hartley, Phil. Trans., vol. 185 (A.), 1894, p. 211. 
+ Hasselberg, ‘‘ Kongl. Svenska Vet.-Akad. Hand.,” B. 24, 1892. 
+ Hemsalech, Ann. der Physik, 2, June, 1900, pp. 331-4. 
§ Loe. cit. 
|| The authors have discovered and directed attention to a convenient source of this metal. The cost 
of extraction will be very much less than by the process employed for extracting gallium from blende. 
s. 
