Band and Line Spectra of the same Metallic Elements. 119 
The spectrum of calcium consisting of the red and green bands and the blue 
line, also the violet lines of potassium, all appear in these spectra, being strongest 
where there is 0°1 gr. Ag, diminishing in intensity down to 0:00001 gr., when 
they are hardly visible. These impurities appear to be caused by dust in the air. 
Experiments were devised for the purpose of ascertaining whether the broad 
diffuse bands are a consequence of the vapour-pressure of the silver in the flame, 
or of its volume, @.e., quantity merely. ‘To this end 0:1 gr. was burnt in the flame 
on one paper, and 0:01 gr. on each of ten papers: the quantity of silver is the 
same in both cases; but the pressure is ten times as great in the former as it is in 
the latter. If the bands are emitted by the silver vapour, and are a consequence 
of the vibrations within the molecule, 0:1 gr., whether introduced all at once, or 
introduced very gradually, should make no difference in the spectrum. If the 
bands are caused by the translatory vibrations of the molecules, then, owing to 
their diminished freedom of motion when the vapour-pressure is increased, the 
bands should be visible when 0:1 gr. is introduced into the flame at one operation, 
but not when the spectrum results from the aggregate effect of ten times 0°01 gr., 
because in the latter case the molecules have greater freedom of motion. In the 
possible third event, where the effect is neither that of the actual quantity of 
silver used, nor simply that of vapour-pressure of the silver in the flame—in other 
words, where it cannot be supposed to be due to either a molecular motion or to 
an intramolecular motion—it must be caused by the large volume of the vapour at 
a considerable pressure, and may possibly be an absorption phenomenon, which 
can be imagined as being produced by a continuous spectrum between \ 4700 and 
4 4290 being emitted by the great body of the vapour, and this having to traverse 
successive layers of vapour at different temperatures, and therefore of relatively 
different densities. 
To determine this point, 0°5 gr. should be placed in the flame on one paper, 
or in one portion ; and also on five papers, or in five portions, each containing 
0-1 gr.; then under the latter conditions the bands should not be more numerous 
with five papers than with one, but they should be more intense. The results 
obtained, which appear to confirm this hypothesis, are as follows :—— 
CHENEY Deserviption of Spectrum. 
0-1 gr. x 5. Bands shown strongly as before. See p. 46. 
O-1 gr. Bands shown feebly but distinctly. 
0°01 gr. x 10. No bands but a continuous spectrum,* which is much stronger 
throughout than the previous spectrum. The two lines at 3383 
and 3280°3 are both very strong. 
0-01 gr. No bands. ‘The two lines are alone visible. 
* Probably due to the filter-paper. 
