^22 ROYAL SOCIETY OF CANADA 



unable to find any periodicity in the motion. Furthermore a change in 

 the amplitude is probably accompanied by changes in the other elements 

 of the orbit which may account for the slight change in the period re- 

 quisite when the Lick observations are brought up to the same epoch 

 as those at Ottawa. 



If all the Ottawa observations are plotted continuously on cross 

 section paper tliey form a curve somewhat similar to the trace given by 

 two beating tuning forks. It shows curves similar to the velocity curve 

 of the figure periodically repeated witli gradually increasing amplitude, 

 then with a sudden diminution followed by another gradual increase. 

 The observations have not been sufficiently continuous or extended 

 enough to decide whether this variation in amplitude is periodic and in 

 any case the very small range combined witli the comparatively poor 

 quality of the spectrum for measurement would render such a deter- 

 mination difficult and uncertain even if a very large number of plates 

 were obtained. 



I have therefore thought it preferable now, as all these successive 

 curves have, so far as can be determined, the same foim, to consider the 

 variations in amplitude as accidental or if you like, as due to errors in 

 measurement; and to obtain a mean curve and from it the elements of 

 the orbit by grouping together into normal places the mean velocities 

 obtained on the 54 nights under discussion. The period chosen was that 

 mentioned above 21.90 days which best suited our own and the Lick 

 observations and which under the conditions cannot probably be im- 

 proved upon. The initial phase To was taken as Julian Day 2,417,96L0. 

 The basis of grouping into the normal places was the phase, the total 

 difference in phase of the nights in a group being kept generally less 

 than half a day, except in three groups where the velocity changes but 

 slowly. 



'These normal places with other data are given in the table below 

 and the places themselves are indicated by circles on the velocity curve 

 in the figure corresponding to the final elements. 



