4 Lt.-Col. J. Waterhouse — On 'Rowland'' s Diffraction Gratings. [Jaw. 



spectra of the first order, and are generally the brightest and purest. They 

 are apparently separated from the direct reflection on one side, and 

 from the spectra of the second order on the other, by dark spaces, but at 

 the violet end a considerable space is filled by the invisible ultra-violet 

 rays extending beyond the H lines, at A 3933, to A 2200 ; a great part of 

 which can be photographed by suitable appliances. Towards the red 

 end also, there is a long series of rays invisible below A, which can also 

 be photographed to about A 10,000, as done by Abney ; and they have 

 been observed to an immense extent further by Mr. S. P. Langley, with 

 his very sensitive thermopile called a bolometer. About the red end of 

 the first order near B, the extreme rays of the ultra-violet of the second 

 order begin to make their appearance when photographed, unless special 

 means are taken to exclude them. Then appear the visible violet rays of 

 the spectra of the second order ; then the indigo, blue, green and so on, 

 till we find the red end of the second order still more overlapped by the 

 violet and blue rays of the third, and so on. With these gratings the 

 spectra of one side are usually more brilliant than those of the other. 

 Every grating has its own peculiarities, due to slight variations in the 

 ruling. With glass gratings four series of spectra are formed round 

 the circumference of a circle, by reflection and transmission ; but with 

 metal gratings only two, by reflection. 



It will be noticeable that the deviation, or angular distance from the 

 centre, of any particular ray in the successive spectra increases as the 

 whole numbers 1, 2, 3, 4. The amount of dispersion in the spectrum of 

 the second order is, therefore, twice that of the first ; in the third order 

 three times, and so on. This is a very valuable property of gratings in 

 enabling spectra to be viewed under different dispersions with the same 

 instrument. It must be noted that, as a rule, in the same way as with 

 prisms, the amount of light decreases as the dispersion increases. 



The amount of dispersion or deviation of any particular ray is 

 inversely proportional to the sum of the width of a polished space and a 

 groove ; the closer and finer the ruling, therefore, the greater dispersive 

 power of the grating. 



Gratings have a further advantage over prisms that the spectra 

 obtained with them are dependent on absolute measurements of the wave- 

 lengths of the different rays. The wave-length may be obtained by 

 multiplying the sum of the width of a transparent interval and a bar, or 

 groove, by the angular deviation of the ray in the spectrum of the first 

 order. 



The deviations of two colours in the same spectrum are proportional 

 to their wave lengths, and this is why in all the spectra the violet is 

 nearest to the central reflection and the red most distant. 



