IN connexion with the spectrum of the sun. 
653 
En outre, tous les metaux que le sodium chasse de leurs combinaisous doivent aussi y 
exister a letat de liberte. 
“ Les nouveaux spectres que j’ai fait connaitre pourront conduire a constater dans 
l’atmosphere solaire la presence du chlore, du brome, de l’iode, du phosphore, etc. 
“ D’un autre cote, de l’absence des raies d’un metal dans le spectre solaire on ne 
saurait conclure celle du metal lui-meme dans l’atmosphere du soleil ; il peut, en effet, 
s’y trouver des metaux, le lithium, par exemple, qui y sont engages dans des combinaisons 
qui ne donnent pas de spectre.” 
In the former communication I have given, I trust, the true explanation of the absence 
of certain lines from the solar spectrum ; and I think in what has gone before there is 
ample evidence that the explanation advanced by Mitsciierlich is absolutely untenable ; 
for at the temperature of the sun, which is high enough to allow hydrogen and even 
sodium to exist uncombined and in a state of incandescence above the photosphere, 
there would be heat enough to dissociate compounds, and therefore to cause the longest 
lines, at all events, of the metalloids to be visible even if they existed in combination as a 
rule, but former observers have recorded* no trace of any metalloids in the solar spectrum. 
As further evidence that there is no chemical combination whatever in the photo- 
sphere, the structure of the spectrum may be also instanced ; it certainly would be very 
different from what it is, did compounds exist in the solar atmosphere ; the least refran- 
gible end of the spectrum would, I hold, be the more, instead of the less complex ; and 
although Professor Young has recently recorded in the spectrum of a sun-spot certain 
appearances which might be imagined to favour the idea of the existence of compounds 
in the comparatively cold downrush into a spot, the general facts, to say the least, seem 
to point the other way, and in all my observations of sun-spots I have never seen any 
thing approaching to the appearance put on by a compound spectrum. 
On this I would also remark that with our present knowledge it is not difficult to 
gather from Father Secchi’s observations on stellar spectra, that if the atmosphere of a 
star contains compound molecules, they at once make themselves very obviously visible. 
Several stars, the spectra of which have been mapped by him, have undoubtedly, in my 
opinion, atmospheres containing compound molecules ; and it may be that the pheno- 
mena of variable stars may be connected with a delicate state of equilibrium in the 
temperature, so that at one time we get the feeble line-absorption of the dissociated, 
and at another the strong band-absorption of the associated elements in their atmo- 
spheresf . Father Secchi’s idea that we have in such stars a prevalence of spot-spectrum J, 
will, I think, not hold ; but this point, which is one of extreme interest, I propose 
to dwell on at greater length in a future communication. In the mean time I may 
remark that I am inclined to attribute spots more to an accumulation of absorbing 
material at a greater pressure than to a similar accumulation at a lower temperature. 
* Angstrom, ‘ Recherches sur le Spectre Solaire,’ p. 37. 
f I find that in this conclusion, drawn from my long and short line observations, I have been anticipated by 
Angstrom, who reasoned from less precise data, op. cit. p. 39.— Note added, January 21, 1874, J. N. L, 
t Secchi, ‘ Le Soleil,’ p. 288 et sep 
