LXXXVI ROYAL SOCIETY OF CANADA 



error ovor that at slit width 0.02Ô mm., but a slit 0.076 mm. gives a 

 marked increase in both accidental and systematic errors. With a three 

 prism spectrograph, dispersion 10 tenth-metres per millimetre at Hy, 

 neither accidental nor systematic errors are much increased by increase 

 of slit width to 0.076 mm. 



Considerable time has been occupied in investigating and perfect- 

 ing the instrumental appliance used in this work, especially the spectro- 

 graph and the correcting lens. The former, which was constructed in 

 the Observatory workshop from designs by Mr. Plaskett, has proved an 

 efficient and accurate instrument, and only some minor improvements in 

 the temperature case and method of applying the comparison spark were 

 needed. 



The correcting lens, whose purpose is to change the form of the 

 colour curve of the objective, bringing Hy instead of A 5600 to a mini- 

 mum focus, enabling a longer range of star spectrum to be photographed 

 in one exposure, was found, after an exhaustive investigation, to give 

 resultant negative aberration of about 2.5 mm. A new correcting lens 

 of larger aperture, designed by Prof. Hastings, gave negative aberra- 

 tion of about 1.5 mm., and it was only after careful reiiguring that the 

 image became practically perfect resulting in a diminution of the re- 

 quired exposure time of about 30^, and in greater freedom from chance 

 of systematic displacement of the lines. 



An auxiliary investigation has been the determination of the wave 

 lengths of the principal lines in the spark spectrum of an alloy of vana- 

 dium and iron, which is used as the comparison source in stellar spectro- 

 scopy. The spectra used in the measures were made by means of a 4-inch 

 concave grating of 10 ft. focus. Owing to the combining of some close 

 vanadium and iron lines, wave lengths obtained from the tables extant 

 are uncertain, and a consistent determination was required to avoid this 

 particular source of error in radial velocity work. 



Micrometric observations of the position angle and distance of 

 double stars as well as of the apparent positions of comets, whenever 

 such are visible, have been carried on with the equatorial telescope, and 

 the scope of the former is being widened. 



The times of immersion and emersion of stars occulted by the 

 moon at Ottawa, have been observed when the weather permitted. 



Photometric observations of some short period variable stars are 

 also being made with a modified form of the Zolhior photomètre attached 

 to the equatorial telescope. 



The shelter for the cœlostat telescope is not yet completed, and 

 work has consequently not been started on spectroscopic observations 



