SPECTROSCOPIC OBSERVATIONS OF THE ROTATION OF THE SUN. 7 
also been closely investigated. The careful mounting and orientation of the instruments, 
it is true, did not warrant the assumption that such an error was to be seriously appre- 
hended. Besides, the arrangement of the observations was such that an error of this 
kind, even though it existed, should practically disappear from the mean values. This 
Fig. d. Ae eicuis showing the Limear rotational velouties observed 
ducing the years 1901-02 and 1905 . 
$--+--++--—+ Oboerov made ev 1901-02 
© © oO » ” ” 1903 
detinn| o o eo °o oO c oO 
eto 6200 Sc SOC 
may be seen from the following consideration. Let the error of the position angle be 
Ap, so that a point on the solar limb whose computed distance from the equator, 
measured in the direction of the position angle, is II, is actually at a heliographic latitude 
II+Ap. If, now, we pass to the opposite side of the pole to a point whose computed 
distance from the equator, again measured in the direction of the position angle, is 
180°—II, such a point would actually be in the heliographic latitude I—Ap. Hence 
TRANS. ROY. SOC, EDIN., VOL. XLI. PART I. (NO. 5). 16 
