212 H. Grew — Period of Rotation of the Sun. 



Since the error due to heating affects the higher latitudes 

 more than the lower, in the ratio of the secants of the lati- 

 tudes ; and since it enters the two series with opposite signs, 

 the coefficients of X, in any accurate work, ought to have 

 opposite signs, always provided there was in reality no very 

 marked change of angular velocity with latitude. 



Combining these two expressions : 



6 = 823' (1 +0-00065 Z°). 



The co-efficient of X is too small to put much stress upon, 

 when we consider the necessarily large errors in the observa- 

 tions from which it was derived. As it stands, however, it 

 indicates that points in latitude 45° rotate in 18 hours less 

 time than points at the equator, while Carrington's expression 

 for sun spots, 



= 865' (I— 0-191, sin * X°) 



would make the rotation period at 45° some 2-| days longer 

 than at the equator. The difference between the two is quite 

 marked. Can it indicate that the spectroscope measures the 

 velocity of regions, corresponding in some degree to upper or 

 lower " trades," the angular velocity of each of which would 

 decrease from the equator to poles ? 



But it is not easy to think of any cause competent to pro- 

 duce " trades " on the sun, since no certain difference of tem- 

 perature* between equator and pole has ever been discovered. 

 The evidence afforded by these observations, therefore, is, that 

 no certain variation of period with latitude has been detected 

 by the spectroscope. 



Wilsing (1. c, p. 436) has examined the motion of faculse 

 with reference to this same phenomenon, but failed to find any 

 " drift " or change of velocity with latitude. 



The following table may, I think, be considered as fairly 

 representing the best determinations of the sidereal period of 

 points on the equator. 



f Hornstdn — from daily range of barometer at Prague [ 9 , , „ 



during the year 1870, \ 



Braun — from daily average of barometer at Singa- \ ,,,,., 



From pore, 1841-1845, \ 



Barometer J Van der Sfok\ — from daily average of barometer, 24*10 



and | " " range ■' " 24 - ll 



Thermometer. " " average " thermometer, 24 - 10 



" " range " " 24'14 



Von Betzold% — from thunderstorms in Bavaria and \ 9i-i9 

 [ Wurtemberg, ) 



* Young's Sun, p. 264. 



f Van der Stok : Natuurkundig Tijdschrift voor Nederlandsche-Indie, Deel 48. 

 The observations extend over 7740 days, were made in the tropics where the 

 amplitude is large, and are entitled to great weight. 

 % Nature, April 11th, 1889. 



