Band and Line Spectra of the same Metallic Elements. 135 
easily boiled away that by the analysis of minerals containing much lead there is 
danger to the operator of lead-poisoning, if the work is conducted in a closed 
photographic room. <A similar risk from poisonous fumes is incurred when 
photographing the spectrum from metallic thallium. These facts are cited in 
order to show that the temperatures at which the banded spectra of leaa, silver, 
cadmium, and thallium were obtained must have been considerably in excess of 
the boiling points of these metals; from which it may be inferred that the similarly 
constituted spectra of antimony, bismuth, tin, zinc, copper, and gold were 
photographed under similar conditions as to temperature, because the quantity of 
material in the flame in each case being smaller, the available heat for vapourizing 
it was in consequence greater. 
The following discussion is restricted, as far as possible, to the chemistry of 
the subject, and, therefore, no reference is made to the beautiful theory of 
J.J. Thomson,* ‘‘On the Structure of the Atom,” nor to the equally interesting 
paper, ‘‘ Kinetic of a System of Particles illustrating the Line and Band 
Spectrum,” + by H. Nagoaka, although neither the facts in detail nor the views 
explained appear to be inconsistent with these physical theories, which account in 
a remarkable manner for homology in spectra and periodicity in the properties 
of the elements. 
On the Relation of the Line to the Band-Speetrum of the same Element. 
It has been shown that in the are under different conditions, the bands of 
compounds such as chlorides,t the bands of elements,§ the arc iines, and the spark 
lines || of the elements may be obtained. Pliicker and Hittorf 4] believed that the 
bands and lines belong to allotropic modification of the same element, the allotropy 
depending solely on temperature. Allotropy is now known to be caused by either 
a loss of energy, as when amorphous carbon becomes successively graphite and 
diamond, or by a gain of energy, as when oxygen becomes ozone; in such 
instances the molecules undergo a change in constitution, but allotropy is unknown 
in a monatomic molecule in the gaseous state. 
The view expressed by Kayser** is, when an element shows two spectra, one of 
lines corresponding to the atom, the other of bands, there is probably in the latter 
case a more complicated structure, au aggregation of atoms, or a molecule. We 
cannot even assume this to be the case when we know the vapour-densities of the 
* Phil. Mag. (6), vol. 7, p. 237, 1904. { Ibid., p. 445. 1904. 
{ Hagenbach. § Pliicker and Hittorf, Basquin, Liveing and Dewar, Hartley. 
|| Ch. Fabry and H. Buisson, C. R., vol. exlvi., April 6th, 1908—‘‘ Presence of Spark-lines in the Arc- 
Spectra of Iron, Nickel, and Cobalt. 
q Phil. Trans., 155, pp. 1-29, 1865. ** Kayser’s ‘‘ Handbuch der Spectroscopie,” vol. ii., p. 258. 
