422 
POPULAR SCIENCE REVIEW. 
render it visible, dissociates the compound according to its volatility ; the 
number of true metallic lines which thus appear is a measure of the disso- 
ciation ; and as the metal lines increase in number the compound bands thin 
out. This principle is applied to the dissociation of the so-called elements, 
especially calcium. Evidence in favour of this is derived first from a series 
of five spectra of this substance at different temperatures, and secondly, 
from the remarkable photograph of the spectrum of a Lyrse obtained by 
Mr. Huggins, in which only one of the solar h lines is represented. 
If the elements be really compounds, these must have been formed at very 
high temperature, to which, by theory, there is no superior limit. If calcium 
be a distinct entity, by working at it alone, we shall never know il whether 
temperature produces a simpler form of the same thing, or whether we 
actually break it up into x + y, because neither x nor y will ever vary. If 
it be a product of lower temperature, then in stars hot enough to enable its 
constituents to exist uncompounded, we may expect them to vary in quan- 
tity, and a Lyrae would thus represent an extreme case.” 
An attempt at Quantitative Analysis by means of the spectrum is the 
next topic introduced, an idea first suggested by Janssen, and demonstrated 
by Frankland, who showed that the spectrum of hydrogen at a pressure of 
20 atmospheres is as continuous as that from burning coal. In the solar 
spectrum the line in the green widens toward the Sun, indicating an approxi- 
mation to a continuous spectrum ; at 20 atmospheres it would be continuous. 
Increase of pressure is therefore the determining power. If so, the same 
lines should become thinner with decreased pressure, and thicken with its 
increase. This they do. There is, therefore, “ individuality ” in the spec- 
tral lines. Experiments by mixture were made to carry out this idea. 
Alloys of silver and gold obtained from the Royal Mint were also employed 
for the purpose, the results being represented by curves, the co-ordinates of 
which showed the composition of the alloys as determined by the ordinary 
method of assay, compared with the varying points of equal brightness 
measured by the micrometer of the observing telescope fitted with movable 
horizontal wires. The parting assay determination, when revised, was 
almost identical with that of the spectroscope. 
A chapter on coincidences of spectral lines, and the elimination of those 
due to impurities, leads to the tenth and last on the elements present in the 
Sun. Kirchhoff and Angstrom’s work is reviewed ; it is shown that when 
the spectra of metallic vapours reversed in the solar spectrum are mapped by 
the new method, the reversed lines are invariably those which are longest in 
the spectrum. It was thus at once seen that zinc, aluminium, and, possibly, 
strontium must be added to the table of solar elements. 
As a preliminary, two lists were formed, one containing solar, the other, 
non-solar elements. The former were found, as a rule, to form stable oxygen 
compounds, and others of this group were then sought for. Cadmium, lead, 
copper, cerium, and uranium were provisionally added to Thalen’s last list. 
As the final result of laborious work, three tables are given: the first, of 
elements whose presence in the Sun’s reversing layer has been confirmed, 
amounting to 22 ; the second, of those whose presence is probable, in num- 
ber 10 ; and the third, of elements absent from the Sun, according to our 
present knowledge. These are carbon, siljcium, and thallium, chlorine, bro- 
mine, and iodine. A-/ t dtrW - 
