120 Hartritry—An Investigation of the Connexion between 
These experiments appear to prove then, first, that the large volume of vapour 
at the atmospheric pressure is necessary to give rise to these particular bands. 
An endeavour to repeat these experiments so as to gain further information 
concerning the composite bands, the heads of which are the lines ) 3383 and 3280°8, 
was rendered futile by experimental difficulties, arising from the large quantity of 
metal which it was necessary to volatilize at one operation before these bands could 
be developed. The results on this account could not be rendered strictly quanti- 
tative, because it was found necessary to maintain an excess of silver vapour in the 
flame, or to place more silver in the flame than could with certainty be vapourized. 
The most interesting fact in connexion with this group of elements is that 
narrow bands have been observed about the same positions, under similar 
conditions, in the spectra of the alkali metals. The manner in which photographs 
of the spectra showing these bands were obtained I have already described in the 
case of lithium, sodium, and potassium.* The foregoing account of the rubidium 
spectrum gives the position of the bands as lying between \ 5100 and 4400. 
The following tabulated statement is useful for comparison :— 
A r At. mass. 
Li, ¢ : . 4677 to 3610 Li, : : : 7:03 
Na, . ¢ . 4800 to 3600 Na, 23°05 
K, : 3 OOMmLO MO O00 K, ¢ c OOM 
Rb, 0 . . 5100 to 4400 Rb, : ; . 85°65 
Ag, . : . 4700 to 4290 Ag, 6 : . 107:93 
With regard to the spectra of the alkali metals it was remarked (loc. c7t.) that 
the bands are individually wider, and also wider apart, the larger the atomic mass 
of the element vapourized, which also means as the density of the vapour increases. 
Silver has the highest vapour-density, and the bands in its spectrum are both 
wider and wider apart than those of the alkali metals; they are produced also 
with very great facility, probably because the silver vapour is not liable to 
oxidation, so that the only change it undergoes is solidification. ‘The bands 
cannot be connected with the atomic weight of the metal excepting through 
its vapour-density. 
The next element to be considered is copper. It has already been pointed out 
that it belongs to the group of metals with monatomic molecules. Its spectrum 
contains two sharp composite bands, and two diffuse bands, between which there 1s 
a close connexion. It will be seen also that the bands are almost the ultimate rays 
of this element. The quantitative results obtained from the copper spectrum lead 
* “On the Thermo-Chemistry of Flame-Spectra at High Temperatures.’”—Proc, Roy: Soc., A, 
vol. 79. 1907, 
