144 MR. CHARLES DE WATTEVILLE ON FLAME SPECTRA. 



close to the point M, the centre of the film, l>oth points M and M, lying on the 

 horizontal line dividing the film into two equal halves. The grating, acting simply 

 as a mirror, forms an image M., of the cross, and, by means of the levelling screws 

 of the grating, M., is brought to the same height as M,. Also, the image of the 

 wires should be perfectly sharp, and this is realised hy sliding the platform carrying 

 the grating either to or away from M, over the clamping screws. Then, when the 

 points M! and M.> are symmetrically situated about the }>oint M, the normal to the 

 grating passes through M and lies in the horizontal plane dividing the film into equal 

 portions. Moreover, the distance of the grating from the film is equal to the radius 

 of curvature of the grating. 



The slit is then illuminated by means of a source of light which gives a line 

 spectrum, and the slit is adjusted in the tube until the lines are seen clearly through 

 the observing eye-piece, which is focussed on the surface with which the film will 

 afterwards coincide. 



The adjustment should then be complete, but this is generally found not to be the 

 case, on account of the different refrangibilities of the eye for the different colours. 

 The further adjustment is then made by the help of photography. The three possible 

 means of adjustment at one's disposal are as follows: (l) The displacement of the 

 slit in its tube this has exactly the same effect for all parts of the spectrum ; (2) the 

 displacement of the grating parallel to itself; and (3) the rotation of the grating 

 about a vertical axis passing through its centre. The last two motions differ from 

 the first in producing effects which vary in magnitude for different parts of the 

 spectrum. 



In this way it is possible to correct the want of sharpness either of the whole 

 spectrum or of any particular portion thereof. By using the three means of adjust- 

 ment methodically, it is possible to secure perfect adjustment after a certain number 

 of trials. A useful plan adopted was to use the grating as a mirror, and by means of 

 the motion on a screen of the image of a small flame to estimate the angle through 

 which the grating was rotated at any time, or, when the grating had only to be 

 displaced parallel to itself, to make sure that there was no rotation about its vertical 

 axis. In this way it was easy to increase or decrease any particular displacement by 

 a desired amount. 



Comparison Spectra. Whenever a flame spectrum has l>een photographed, a 

 comparison spectrum has. been photographed with it, this being either the spark or 

 the arc spectrum of the metal under investigation. It was important, in order to 

 avoid any accidental relative displacements of the lines, that the burner and the 

 electric arc or spark should lie exactly in a straight line with the centre of the 

 grating. This condition was realised in the following manner; The electrodes 

 E, E' are adjusted so that the gap between them is at the same height as the centre 

 of the grating (and therefore at the same height as the centre of the photographic 

 film). The slit F then being opened widely and the spark set going between E and 



