DECEMBER 1, 1899.] 
with the dark lines in the solar spectrum 
ascribed by Rowland to carbon. On August 
17th the yellow carbon fluting, which is 
more difficult than the green fluting, was 
also found. These observations reveal an 
interesting similarity of the sun to red stars 
of Secchi’s fourth type, in which a dense 
absorbing atmosphere of carbon (which is 
far more conspicuous than in the sun) has 
recently been found with the 40-inch tele- 
scope to be surmounted by an atmosphere 
giving a spectrum of bright lines. 
Some New Forms of Spectroheliographs. 
Of the various forms of spectrohelio- 
graphs described in my previous papers, the 
simplest and best is undoubtedly that in 
which the instrument is moved as a whole 
at right angles to the axis of the telescope, 
the solar image and photographic plate re- 
maining stationary. It is not always pos- 
sible, however, to employ a spectrohelio- 
graph of this form. With the forty-inch 
telescope, for example, the motion of the 
very heavy spectroheliograph required could 
not be accomplished without jarring the in- 
strument. For this reason it has been de- 
‘cided to cause the solar image to move 
across the first slit by means of the slow 
motion declination motor. The first and 
second slits are fixed with reference to 
each other, and the photographic plate is 
moved across the second slit by means of 
a screw driven by the same motor, which is 
mounted on the tube of the forty-inch tele- 
scope. A wide range of exposures can be 
secured by means of a system of change 
gears. This spectroheliograph, which has 
an aperture of 64 inches, is now nearly 
ready for trial. 
Two other forms of spectroheliographs 
may occasionally prove useful. The solar 
image is moved across the first slit in the 
one case by means of a photographic doub- 
let, of large field, mounted between the slit 
and the principal focal plane of the image 
SCIENCE. 
795 
lens, and in the second case by means of a 
right angle prism, placed immediately in 
front of the first slit, with hypothenuse face 
parallel to the optical axis of the colli- 
mator. A suitable combination of mirrors 
may be used instead of the prism. The 
doublet or prism are connected with a car- 
riage bearing the photographic plate across 
the second slit, and are moved in a direc- 
tion at right angles to the optical axis of 
the collimator. Hither device, used in con- 
junction with a heliostat, affords an easy 
means of transforming a large fixed labora- 
tory spectroscope, of almost any type, into 
a spectroheliograph. 
Comparison of Stellar Spectra of the Third and 
Fourth Types. 
As previous observations of the faint red 
stars of Secchi’s fourth type have shown 
none of the lines in their spectra it has 
been impossible to effect any satisfactory 
comparison of these stars with the red 
stars of the third type. So far as their 
characteristic features go, the spectra are 
quite dissimilar, the pronounced carbon ab- 
sorption bands of the fourth type having 
no counterpart in the banded spectra of 
the third type stars. The photographs of 
these spectra recently obtained with the 
forty-inch telescope, by Mr. Ellerman and 
the writer, contain hundreds of lines, and 
render comparisons possible. From these 
plates it has been found that in certain 
limited regions third and fourth type 
spectra are almost identical. It is, there- 
fore, probable that the stars of these two 
great classes are closely related to each 
other and to stars likethe Sun. Thestudy 
of fourth type stars is being continued at 
the Yerkes Observatory. 
(To be continued. ) 
Epwin B. Frost, 
Acting Secretary. 
YERKES OBSERVATORY, 
WILLIAMS BAY, WIS. 
