102 DR J. HALM ON 
reach of our large astrophysical observatories, and there are several beautiful helio- 
meters which might be resuscitated for this purpose from their present unprofitable state 
of repose. Nor is the present the only investigation which can be taken up with an 
instrument of the kind here described. ‘The construction of the apparatus, combined 
with its high dispersive power, makes it possible to separate by a mere glance the 
telluric lines from the solar ones, and at the same time to determine by a single 
observation the wave-length of a line with an accuracy of about +0°05 tenth-metres. 
The instrument appears therefore to be particularly adapted for an investigation of the 
telluric spectrum in accordance with the method first suggested by Cornu. 
As regards the accuracy of the determination of the rotational velocities, I am 
confident that, under more favourable atmospheric conditions, the annual output of 
observations can be considerably increased, and thereby the probable error of the 
annual means correspondingly lessened. It does not seem to me impossible that in this 
way changes of only one to two hundredths of a kilometre could be traced with certainty. 
I must not conclude this paper, however, without drawing attention to an incon- 
venience encountered in this spectroscopic determination of the rotational velocity. 
With a hazy atmosphere, a not inconsiderable quantity of scattered daylight, reflected 
from the particles of the aqueous vapour contained in the air, is thrown upon the slit of 
the spectroscope. As a certain percentage of this light will have emanated from 
the interior of the sun’s disc, the solar lines of this ‘‘day” spectrum appear less dis- 
placed than the lines of the two solar limbs. By a superposition of the two spectra, 
the intensity of the lines will therefore be lessened on the one side and increased on the 
other, the effect always being to bring the lines nearer to their normal position. The 
velocities obtained under such conditions must therefore be too small. I have often had 
occasion to observe this phenomenon when, after a measurement made under favourable 
atmospheric conditions, the sky was suddenly overspread with a veil of haze, a meteoro- 
logical feature not at all infrequent in the climate of Edinburgh. When the observation 
was then repeated, the measured displacement was invariably found less than that 
obtained before. Observations under misty conditions of the sky were therefore care- 
fully avoided. A reliable scale for the transparency of the atmosphere was supplied by 
the optical appearance of the dark band separating the spectra of the two limbs in the 
viewing telescope. Whenever this space became so bright as to show traces of the 
absorption lines of the solar spectrum, the observation was broken off. All the measure- 
ments used in this discussion were made at moments when the band mentioned was 
uniformly dark, so that the spectrum of the diffused daylight cannot have seriously 
influenced the measured displacements. 
