Oct. 27, 1870] 
in the three atmospheres :— 
~ La oc 
‘ 
\ 
NATURE 
BIE, 
cases, the sp. gr. of the enclosed gas masses must not be taken 
as higher than that of the enclosing liquid, since otherwise the 
compressed gas masses would, in virtue of the Archimedean 
principle, sink down into the interior of the sun. 
The sp. gr. of the sun is, poe to the latest determinations, 
1°46. 
Substituting this value for o and for a (in Formula v.), the 
above found value 40,690, also for # the value 8” in metres, we 
obtain for the extreme values A, = 0050 m. above given, the 
following values of ¢: 
for fj, = 0.500m. - 7% 
; for J, = 07050m. ¢ 
therefore in mean, ¢ = 27,700". 
é On differentiating Equation (5) by ¢, the differential quotient 
do 
29,500° 
26,000°, 
is negative. From this follows that the values found above 
for are also minimum values. 
From the mean value of ¢ for the temperature of the sun’s 
atmosphere the value of /, is found = 0 180m. These values 
will be those made use of in the following calculations. 
It may be noticed in connection with the high numbers ob- 
tained for the temperature values, that they are about eight 
times higher than the temperatures of combustion of a mixture 
of detonating gas as found by Bunsen, and that iron must per- 
manently exist in a gaseous condition in the sun’s atmesphere. 
With the above value for ¢ = 27,700° we obtain from Formula 
(1.) for the inner temperature 
4 = 638,4c0 
substituting these two values of ¢ and ¢, in Formula (11.) we have 
a 22°T 
2. 
é.é., the pressure in the interior of the space from which the pro- 
tuberances break forth is 22"1 times greater than the pressure on 
the surface of the liquid separating layer. Further substituting 
the value for ¢ in Formula (Iv.) and assuming as before the value 
of & to be 8”, we have 
Pa 
Pr 
as the relation of the pressure on the fluid surface of the sun to 
the pressure at the height 4, where the hydrogen spectrum, in 
consequence of the pressure, begins to become continuous. 
Substituting for J, the above value of 0180 m, mercury, we have 
Pa 184,000 atmospheres, 
and consequently for 2; = 4,070,000 i. 
If the depth be calculated at which in the interior of the 
liquid mass of the sun which has a sp. gr. of 1°46, and simply 
as the result of the hydrostatic pressure, this maximum pressure 
of #, would be attained, it is found that this would occur at a 
depth of 139 geographical miles below the surface, £4, ata depth 
of about 1°46 arc seconds, or stg of the sun’s semidiameter. 
Even if the liquid condition be put quite out of question, 
= 766,000 
and, under assumption of a much larger atmospheric envelope 
of hydrogen, the depth in it be calculated, at which the atmo- 
spheric pressure becomes equal to the inner pressure ;, it is 
found that even assuming a temperature of 68, 400°, that depth is 
only 27" below the visible edge of the sun’s disc, or about 34 of 
the sun’s apparent semidiameter. 
This circumstance shows how rapidly the pressure must in- 
crease towards the interior of the sun’s body, and thus justifies 
the assumption that in the interior of the sun, even at such high 
temperatures, the permanent gases, for example, hydrogen, can 
only exist in an incandescent liquid condition. 
(5 
A surprising result is obtcined if, under the assumpt’on of a 
_ nitrogen or oxygen atmosphere of equal weight and temperature to 
the hydrogen atmosphere above considered, the pressure be cal- 
cwlated which is reached in those atmospheres at heights at which 
the hydrogen spectrum commences to become continuous. If at 
a depth of 8" below the visible edge of the sun’s disc, #.é@, a 
the riwean of the supposed separating layer, the pressure of the 
three atmospheres of hydrogen, oxygen, and nitrogen be assumed 
as equal, and that fa 184,000 atmospheres, a value which 
from the above, corresponds to the assumed value of #,. The 
following values are obtained for the pressures at the temperature 
above found ¢ = 27,700° on the surface of the sun's vész/e disc2 
Hydrogen #, = 180 millimetres. 
‘Nitrogen 2, = 323r075 
It follows ‘rom these that, the assumptions made, the quantities 
of the two latter gases are, in proportion to the quantity of hy- 
drogen in that layer in which the spectrum of the latter com- 
mences to be continuous, infinitely small. This would, as is 
evident, also be the case if the weights of the two atmospheres 
were assumed to be many million times greater, although having 
regard to the specific gravity, a 14-times smaller weight of nitro- 
gen and a 16-times smaller weight of oxygen would suffice, in 
order that under the assumed conditions the density of these two 
gases should coincide at the base with that of hydrogen, 
According to our former considerations, the sun’s mean speci- 
fic gravity would also in this case have to be assumed as the 
maximum value of the density at the base of these atmospheres, 
and it is easy to calculate, with the help of Formula 3, and the 
known specific gravities of oxygen and nitrogen, how high the 
weights of these two atmospheres would have to be assumed in 
order to attain this maximum value. 
As result is obtained, that the weight of the oxygen atmosphere 
could only amount to °56, that of the nitrogen atmosphere to “64 
of the weight of the existing hydrogen atmosphere. 
If therefore the simultaneous existence of these three gases on 
the sun’s surface he assumed, and the influence of atmospheric 
motion be disregarded, the rays emitted by the continuous spec- 
trum of the hydrogen layers would, on their path to our eyes, 
pass through so small a number of incandescent nitrogen and 
| oxygen particles, that the absorption caused thereby is a vanish- 
ing one, and therefore, as is in fact che case, the presence of 
oxygen and nitrogen in the sun’s spectrum could not be demon- 
strable by dark lines. 
Although the motion of the gases is active in lessening the 
differences just considered, the existence of the chromosphere 
proves clearly the slight influence of this action in consequence 
of the great intensity of gravity, and the considerable height of 
the layer cousilered (ccmpare Formula 4). 
In order, however to explain through the circumstance indi- 
cated the absence of lines in the sun’s spectrum of two bodies of 
| such universal distribution on the earth as nitrogen and oxygen, 
the very slight emissive power of the permanent gases in pro- 
portion to that of volatilised bodies must also be taken into con- 
sideration. If the emissive power of different gases at the same 
temperature for rays of the same refrangibility be referred to 
equal very minute weights of these gases,* the before-quoted 
experiment of Wiillner’s, in which the small amount of sodium 
yolatilised in the Geissler’s tube emitted more light than the hy- 
drogen gas under a pres-ure of 1,000mm., gives a beautiful prou 
of the extraordinary difference of emissive—and consequently 
according to the theorem of Kirchhoff of absorptive—power of 
different gases at the same temperature. Only by consid ration 
of this circumstance is the contradiction removed, which could 
be deduced against the above explanation of the absence of 
nitrogen and oxygen lines from the fact that in the sun’s spectrum 
the lines of bodies are present whose vapour densities, as a con- 
sequence of their simple relation to the atomic weights, must be 
much higher than the density of oxygen and nitrogen. 
From these considererations, partly directly and partly indirectly 
through a longer series of conclusions, a detailed exposition of 
which I reserve for another occasion, the following result :— 
I. From the absence of lines in the spectrum of a star shining 
in its own light, the absence of the corresponding element must 
not be inferred. 
2. The layer in which the reversal of the spectrum occurs is 
different for each element—the higher the vapour density, and 
the lower the emissive power of the element, the nearer it is 
situated to the centre of the star. 
3. For different stars, under otherwise similar conditions, this 
layer lies the nearer to the centre the greater the intensity of 
gravity. 
4. The distances of the reversing layers of the separate ele- 
ments, both from the centre of the star and from each other, 
incrcase with an increase of temperature. 
5. The spectra of different stars are under otherwise similar 
Conditions the more rich in lines, the lower their temperature 
and the greater their mass. 
6. The great difference of intensity of the dark lines in the 
sun’s spectrum and other fixed stars does not depend only on the 
differences of absorptive power, but also on the depths at which 
the reversal of the respective spectra takes place. 
Ja conclusion, I would offer a few remarks on the application 
of the observations carried out on rarefied gases to the heavenly 
bodies. Lecoq de Boisbandrau* has recently pointed out, with 
* Compt. Rend. Ixx. p. rogr. 
