ASTRONOMY: W. 5. ADAMS 
157 
appear that a relationship between certain spectral characteristics and 
the distances of stars could hardly be credible, since it would appear 
like a correlation between two utterly unrelated subjects except in so 
far as the scattering of light in space might connect them. In fact, of 
course, it is not the distances but the absolute magnitudes of stars which 
have an influence on the character of the spectrum Hnes and such an 
effect, far from being improbable, is rather to be expected than not. 
The derivation of the distances is merely a by-product resulting from 
the combination of real, or absolute, with apparent magnitudes. 
An important gain in the value of this method of determining stellar 
magnitudes and distances should result from an increase in the number 
of measured parallaxes of bright stars of small proper motion. Such 
stars will on the average prove to be very luminous, and, as already 
stated, the portion of the curves connecting Hne intensity with absolute 
magnitude is subject to much more uncertainty in the case of the high 
luminosity stars than in any of the others. It is probable that after 
such a revision has been made the method will find its most important 
application as a means of distinguishing these giant stars in the stellar 
system. 
INVESTIGATIONS IN STELLAR SPECTROSCOPY. IV. SPECTRO- 
SCOPIC EVIDENCE FOR THE EXISTENCE OF TWO 
CLASSES OF M TYPE STARS 
By Walter S. Adams 
MOUNT WILSON SOLAR OBSERVATORY. CARNEGIE INSTITUTION OF WASHINGTON 
Received by the Academy, February 8, 1916 
The principal distinguishing feature of the M type of stellar spectrum 
on the Harvard system of classification is the presence of absorption 
bands due to titanium oxide. These bands increase in intensity for the 
successive subdivisions Ma, Mb, and Mc. The star a Orionis, in which 
they are present in moderate intensity, is selected as a typical Ma star 
by the Harvard observers. Since these bands may be seen faintly in 
stars of the K5 type of spectrum it is necessarily largely a matter of 
judgment whether in any given spectrum they are sufiiciently strong to 
warrant classifying the star as Ma, or whether it should still be retained 
within the K type. 
For types of spectra previous to M the principal basis of classification 
is the intensity of the hydrogen hnes. These reach a maximum in the 
A type, and grow fainter in the successive types F, G, and K. Of the 
hydrogen Hnes in a Orionis, however, Miss Cannon, in the course of 
