4 - s. Hirayama. 



may observe that there is a tendency in the spot-groups to concent- 

 rate upon one particular period, but each curve is not symmetrical 

 about the ordinate of maximum number. Another important 

 point is the existence of a secondary concentration. At first I 

 thought that if there existed a gi-eater numljcr of recorded 

 observations, then I could get a sort of probability curve, but 

 actually it seems likely that each curve will be a combination of two 

 probability curves. Because of the impossibility of determining 

 rigorously the values of several constants Avith non-linear equations 

 by the method of least squares, and moreover because of the 

 doubtful nature of the problem, I have endeavoured to determine 

 the positions of the principal and secondary maxima by graphical 

 method. When the maximum is not well pronounced, I have 

 derived the smoothed curves by compounding two sjaiimetrical 

 curves about the directions marked I and II. All doubtful cases 

 (marked? on the plates) have been excluded. I have thus obtained 

 the following results: 



Taele II. 



Angular Velocity corresponding to Maximum I. 



