146 



THE ROYAL SOCIETY OF CANADA 



and hot readings would give the effect due to vapour alone, within the 

 probable error of readings. 



In October of last year we commenced to repeat the experiment 

 with Thallium, working at a higher temperature, about 800°C., and 

 using a more accurate method of measuring the small deflections. 

 This time we also expected to be able to determine the absolute 

 refractivity. Great difficulty was experienced in obtaining a steady 

 deflection at a higher temperature, and in repeating readings, either 

 hot or cold, on successive days. However, values were finally 

 decided upon, giving a dispersion curve of a somewhat similar shape 

 to that previously obtained, but an absolute refractivity very different 

 from our former estimate. 



We next attempted to perform the experiment with Calcium, for 

 which we expected to find a deflection just reasonably measurable, 

 but obtained readings comparable with those for Thallium. It was 

 now realised that the air in the ends of the tube might have a con- 

 siderable effect on the deflection, and an experiment was commenced 

 to determine the necessary correction. A carefully cleaned cell was 

 highly exhausted, and the experiment performed with the empty cell 

 under as nearly as possible the same conditions as before. The read- 

 ings obtained for this correction were rather irregular, and were 

 sometimes more, sometimes less, than those for Calcium, while they 

 amounted to 60 to 75 per cent, of those for Thallium. It was obvious 

 that no value could be attached to the corrected readings, even for 

 Thallium, and a new tube was made with the object of eliminating the 

 end effect- 



A plate of quartz was chosen which deflected a beam of light not 

 more than 3 or 4", and which made no noticeable change in the focus. 

 From this four plates were cut and sealed into a tube, as shown in 

 Fig. 2. The intention was to evacuate the two end spaces and to 



va eu u?n 



O 



Fig. 2 



place the metal in the centre space, but as a preliminary test all three 

 spaces were exhausted. 



It was found that after the plates were sealed in, the cold cell 

 deflected the light about 50". and the focus of the examining telescope 



