Coloured Flames of High Luminosity. 245 



with the exception o£ the first, are disconnected by closing 

 the pinch-cocks. A very gentle current of air is now 

 passed through the system and coal gas admitted until 

 a flame is obtained which is still luminous near the tip. 

 Oxygen is then slowly added until the flame has become 

 non-luminous and the cone, on the other hand, very 

 bright. The flame may be from 12 to 18 inches long 

 and not more than \ inch in diameter. The cone should be 

 not more than 1 inch long. The flame is now coloured by 

 passing sparks in the sprayer in use. To proceed to the 

 next sprayer pinch-cock P 2 is opened. After this, and only 

 after this, pinch-cock P 1 of the first sprayer is closed. The 

 air will now be passing through the second sprayer. Next, 

 the negative electrode of this sprayer is connected to the 

 leyden-jar and sparks passed as before. In this way the 

 flame can be fed in succession with the sprays from six 

 different solutions without necessitating any readjustment 

 of gas mixture. 



The flames obtained in this manner are exceedingly well 

 suited for showing, in a large lecture theatre, the spectra of 

 the more volatile elements, such as Ca, Sr, K, Cu, &c, to an 

 audience provided with small replica transmission gratings. 

 The relatively great length and thinness of the flame obviates 

 the necessity for a spectroscope slit. The light from the flame 

 is simply allowed to fall on the gratings at a smaller or greater 

 angle of incidence according to the brightness or purity of 

 spectrum desired. Small replica gratings suitable for this 

 purpose can easily be prepared in large quantities as 

 follows: — A cast is mnde from a larger grating according 

 to Wallace's method * and provisionally mounted face down 

 on a piece of plate-glass. By means of a sharp knife (a 

 Gilette razor-blade does well) the ruled portion of the film 

 is cut up into small rectangles of, say, not more than 

 •J by 1 inch sides. These small gratings are then floated 

 off under water and remounted face down in the usual way 

 on small slips of glass. They should be provided with a 

 cover-glass and edged with black paper, like lantern-slides. 

 It is not absolutely necessary to cement them in. 



The very brilliant sodium flame yielded by this method 

 will be found invaluable in interference experiments, and 

 the holes in the burner may, if required, be arranged in a 

 single row, so as to produce a broad sheet of flame. 



July 10, 1917. 



* R. J.Wallace, ' Astrophysical Journal.' vol. xxii. p. 123 (1905), and 

 vol. xxiii. p. 96 (1906). 



