24 THE ROYAL SOCIETY OF CANADA 
(C) The latitudes to be observed in the special region are—0°, 15°, 
30°, 45°, 60°, 75° and if possible 80° and 85°. The latitudes to be 
observed in the general region are—0°, 30° and 60°. 
(D) 15 or 20 lines are to be measured in the special regions, these 
to be selected to include as many elements as possible especially those 
of high or low atomic weight; about 10 lines, selected by the Secretary 
of the Committee after consultation, are to be measured in the general 
region. 
3. The principal objects of a study of the sun’s rotation by the 
spectroscopie method are :— 
(a) The accurate determination of the velocity of rotation at 
various latitudes and the derivation of a formula representing the 
variation of velocity with latitude. 
(b) A definite conclusion in regard to the existence of variations 
in the rate of rotation. 
(c) The investigation of the rate of rotation, as shown by the 
lines of different elements and of the arc and enhanced lines of the 
same element, to determine whether either the absolute rate of rotation 
or the law of variation with latitude difiers for different elements. 
(d) The detection of possible systematic proper motions or drifts 
in the sun’s reversing layer. 
4. In accordance with the above plan three series of plates were 
made during 1911, two in the special region at 45600 and one in the 
general region at 44250. With a solar diameter of, on the average, 
227 mm., the distance of the observed points from the limb in the 
first series, at 45600, varied from 3.0-4.5 mm.; in the second series, 
also at 45600, was nearly 10 mm.; and in the third series at 44250 
was about 6.5 mm. The distance was varied in order to see if any 
difference in the rotational value was obtained, and also to see if much 
change in the definition occurred as the distance from the limb was 
increased. As will be seen later, the difference, if any, is slight both in 
the velocity and the definition. Owing to the considerably larger cor- 
rections required to reduce the measured to the actual values of the 
rotation as the distance from the limb increases, it is not deemed 
desirable to, in future, make the spectra from points at a greater distance 
than 5 mm. from the limb. 
PRECAUTIONS. 
5. In all these plates particular care was taken to guard against 
every known cause of instrumental and other error tending to intro- 
duce spurious displacements of the lines, and the experience of one 
of the writers in stellar radial velocity determinations was of great value 
in this similar work. Temperature changes and flexure, the chief 
