4 
- Sherman—Study upon the Atmosphere of B Lyre. 129 
of increasing hydro-carbon, and of an almost entire absence of 
carbonic acid. The downward slope to secondary minimum 
corresponds to a period abundant in carbonic acid, almost want- 
ing in oxygen with variable evidences of hydro-carbon. The 
remaining portion of the curve corresponds to a period in which 
the evidences of hvdro-carbon, oxygen, and carbonic acid are 
irregularly variable; as would be the ease if the light waxed 
and waned by a series of undulations similar to those which 
Chandler* tells us are actually observed. The final slope cor- 
responds to a period of decreasing carbonic acid, increasing 
hydro-carbon, increasing oxygen, preparing as it were for a 
conflagration at primary minimum, and such apparently takes 
place. It is of interest to remark that hydro-earbon probably 
formed a part of the spectra of both Nova Cygni and the star 
near y, Orionis, but was only traceable near the star’s maximum. 
Recalling Lockyer’s result in the study of the solar atmos- 
phere, agreeing as it does with certain observations which we 
have quoted above, we may picture to ourseives the condition 
of the stellar atmosphere, and the action therein somewhat as 
follows: An outer layer of hydrogen positively electrified, an 
inner layer of oxygen negatively electrified, and between them. 
a layer of carbon mingling on its edge with hydrogen. The 
electric spark passing through the mixture forms the hydro- 
carbon compound, whose molecular weight carries it into the 
oxygen region where combustion ensues with the formation of 
carbonic acid and aqueous vapor, both of which descending 
under the influence of their molecular weight are again dis- 
sociated by internal heat, and return to their original positions. 
Under the insight which this result gives we have found the 
spectrum of the nebuleet referable to low excitation hydrogen, 
the spectra of the bright line stars referable to high excitations 
oxygen and hydrogen of higher or lower excitation according 
as the central star is of a high or low magnitude, and as far 
as the accuracy of the observations permits, t Coron, Nova 
Andromedx, Nova Cygni and the star near y, Orionis itself a 
variable, likewise referable to the same spectra similarly con- 
ditioned. There is also reason for thinking that a similar 
atmosphere in similar physical conditions, lies between us and 
the sun, and it seems as if we might consider that from the 
faintest nebula to the most highly finished star we have but 
progressive stages of the phenomenon here presented. ‘T'o 
observe, to differentiate, to classify, to interpret, presents a vast 
field of deep interest and high promise. This line of investi- 
gation, however, will appreciate both new work in the labora- 
tory and new methods with the telescope. To prepare these is 
now our charge; later we hope, under some clearer sky, to 
reénter the field. 
*Chandler. Astronomical Journal, vol. vii, No. 1. + Ibid., vol. vii, No. 6. 
