1028 M. L. Dunover and Prof. Wood on Photometr 



IG 



the bulb. The losver capillary is then sealed, and finally the 

 upper. The pump should be working vigorously all the 

 while, as the brilliancy of the resonance depends upon having 

 the highest possible vacuum. In our experiments we heated 

 the bulb for twenty minutes, to make sure of getting rid of 

 all of the gases, and the sodium was previously heated f//< 

 vacuo, but these extreme precautions are not necessary in 

 the preparation of bulbs for lecture purposes. 



The bulb was supported by a wire in a column of hot air 

 rising from a large tube of fireclay with a large Meker 

 burner at the bottom, by means of which a fairly uniform 

 temperature up to 400° could be obtained. For lecture 

 purposes it is sufficient to heat the bulb by waving a large 

 Bunsen flame over its entire surface. The arrangement of 

 our apparatus is shown in fig. 2. As a standard source of 



Fijr. 2. 



comparison we used a sodium flame A behind a piece 

 of ground glass mounted behind a pair of large Nicol prisms 

 B and C. The sodium bulb D was first coated with magne- 

 sium oxide, by holding it above a piece of burning Mg ribbon. 

 It was then wiped clean, a small square patch of the oxide 

 being left on one side. The image of the window in the 

 iron chimney surrounding the exciting sodium flame F was 

 formed upon the square of magnesia and the adjacent clear 

 glass of the bulb. It was our object to measure the ratio of 

 the intensity of the magnesia and the vapour of sodium 

 under equal illumination. This was done by means of a very 



