TRANSACTIONS OF SECTION A. 4.65 
3. A Mountain Observatory in South India. By ©. Micum Smurvu. 
4. The Variability in Light of Mira Ceti and the Temperature 
of Sunspots. By Rev. A. L. Corriz, S.J., F.R.A.S. 
The existence of a spectrum of bands in sun-spots has been recorded by 
dbservers since 1869, when they were first observed by Secchi. A fluted or 
banded spectrum is characteristic of chemical compounds. Hence the view enun- 
ciated by Young in 1872 that a fall of temperature over sun-spots allowed the 
formation of such compounds, Recent spectroscopic laboratory and observatory 
work at Mount Wilson and South Kensington has strengthened the probability of 
‘this view. Some of the bands in the red end of the spectrum are due, as Hale 
and Adams have shown, to titanium oxide, others in the green have been recently 
attributed to magnesium hydride by Fowler. The purport of the present paper is 
to adduce an argument in favour of the relatively low temperature of sun-spots, 
from the behaviour of the bands of titanium oxide in 0 Ceti, when the star is at 
two different temperature levels, represented by a whole magnitude in luminous 
power. At the maximum in 1897 the magnitude of the star was 3:0, at that of 
December 1906 it reached the brilliancy 20 mag. Two series of plates of the 
spectrum of the star were taken under precisely identical conditions of instrument 
and plate by Father Sidgreaves at Stonyhurst at the two periods. In 1897 
fourteen bands were photographed between H, and H,: these bands are much 
reduced in intensity in 1906, those with heads at \ 4757, 4803, and 4842 having 
entirely lost their winged appearance. Concomitant evidence of the greater tem- 
eee of the star in 1906 is furnished by the behaviour of the hydrogen lines. 
ot only is the whole series from H, to H, present, with the exception of H,, 
covered by calcium absorption, but H, and H, are winged, presenting somewhat 
of the appearance of the hydrogen lines in the new stars Nova Aurige and Nova 
Persei. Again, there is a gradual darkening of the titanium-oxide bands in the 
spectra of stars from a solar star such as a Tauri of Type II, through 8 Andro- 
medze, a Orionis, y Geminorum, 8 Pegasi, to a Herculis and o Ceti; and in the 
transition from Type II. to Type [I1. Mira Ceti is further removed from the solar 
type than a Orionis or a Herculis, and the characteristic sun-spot lines of titanium 
and vanadium are more prominent in its line spectrum than even in the case of 
a Orionis. These facts furnish further links of similarity between the sun-spot 
spectrum and that of o Ceti. Hence, to sum up, the undoubted presence of the 
chief constituents of the line spectrum of sun-spots as intensified in stars of 
Type III.; the presence of the bands of titanium oxide, also recognised in sun- 
spots; the partial disappearance of some of these bands and the totai disappearance 
of others when Mira Ceti attained a greater luminosity at maximum than is 
usual, and this too enhanced by a bebaviour in the hydrogen lines akin to that 
observed in new stars; the substitution of a line for a banded spectrum in a series 
of stars on an ascending scale of temperature—these are all phenomena which 
when linked together point to the conclusion that the temperature of sun-spots is 
lower than that of the solar photosphere. 
5. On a Method of improving the Constants of the Plates for the Astro- 
graphic Catalogue. By Professor H. H. Turner, F.RS. 
The centres of the plates for the Astrographic Catalogue in any one zone are 
at intervals of eight minutes apart near the equator and more elsewhere. Suppose 
we point the telescope to one of them and give an exposure of four minutes. 
Without disturbing the plate at all, we have now four minutes in which to set for 
the guiding star of the next region, pick it up, and commence (and finish) an 
exposure of four minutes to this new region at exactly the same hour-angle as 
before. Then we again have four minutes to pick up the next region, and the 
process may be continued indefinitely. When the plate is developed, a réseau 
1907. HH 
