58 THE ROYAL SOCIETY OF CANADA 
The agreement of Dunér’s, Adams’, 1908, and the Ottawa values, 
except for small and nearly constant differences, is quite striking, and 
gives good grounds for the belief that the law of variation with latitude 
is represented to a high degree of accuracy by a Faye formula with co- 
efficients approximately the same as those given in these three formule. 
23. In regard to the absolute value of the rotational velocity 
the question can not be regarded as by any means settled. Considering 
the velocity values at the lower latitudes we find that Halm and 
Adams get nearly the same values, Dunér and Storey and Wilson are 
about 1 per cent. higher, Plaskett about 2 per cent. lower and De Lury 
about 4 per cent. lower. But at the higher latitudes Dunér and Adams 
(1908) agree, Plaskett is 2 per cent. lower as before, De Lury about 5 
per cent. lower, Storey and Wilson are 5 per cent. higher, while Halm 
and Adams (1906-7) are some 15 or 20 per cent. higher. At the 
equator Plaskett’s values are in practical agreement with the motion 
of sun spots. As it is generally considered that the reversing layer 
and sunspots are at the same level from the practical identity of their 
spectra, this so far as it goes gives weight to the lower value of 14°.4 
at the equator. On the other hand as the latitude increases the sun 
spot velocities agree better with the higher values of the reversing 
layer such as those of Halm and of Adams’, 1906-7 observations. 
24. These differences in values may be due to one or more of 
three causes:—a. A variation in the rate of rotation of the Sun. 0b. In- 
strumental errors. c. Personal errors of measurement. 
(a) The question of a change in the rotational velocity of the Sun, 
which was raised by Halm *, was quite fully discussed by Adams,f 
who reached the conclusion that the evidence to date was against 
variation. The later values by Storey and Wilson and those obtained 
here, of which the former is higher and the latter lower than Adams’ 
results, would indicate a variation in the rate of rotation were it not 
for the possibility of small instrumental and the probability of personal 
measurement errors (Sections 15-17). As it is, until the latter are 
eliminated, it will be impossible to make any definite statement in 
regard to either the variation or constancy of the rate. Certainly the 
possibility of a variation must, until further evidence is available, be 
taken into account in considering the difference obtained. 
(b) So far as instrumental errors are eoncerned although every 
known precaution was taken to avoid them, it is possible that some 
small systematic effects may be present in these results. The only 


* AN. 173 p. 294. 
+ Adams & Lasby, p. 115. 
