﻿Gratings with Controlled Groove Form. 311 



when the direction of the electric vector was perpendicular 

 to the direction of the groove, and vanished when these 

 directions were parallel. 



In the present paper I shall describe a still more remark- 

 able grating which shows a bright line in the continuous 

 spectrum of a white source, which by actual measurement 

 is only 3 Angstrom units in width, i. e. of a width equal to 

 one half of the distance between the D lines. As the angle 

 of incidence is varied the position of the bright line in the 

 spectrum changes, and, what is of especial interest, its wave- 

 length corresponds exactly to the wave-length which is just 

 passing off from the grating at grazing emergence in the 

 4th order spectrum. This is in agreement with the surmise 

 expressed by Lord Rayleigh (Phil. Mag. July 1907), that 

 abnormalities are to be expected at such angles of incidence 

 that spectra of higher orders are just passing off. The 

 bright line in question appears in the first order spectrum 

 on the opposite side from that on which the grating concen- 

 trates litjht, and is about as bright as the continuous spectrum 

 upon which it appears projected, so that the brilliancy at the 

 point in question is about double that of the rest of the 

 spectrum. In viewing the spectrum of the arc, the bright 

 line resembles the bright lines which sometimes flash out 

 upon the continuous spectrum of the crater. 



As I showed in my paper upon the echelette grating (this 

 Journal, Oct. 1910), if a grating is ruled upon a polished 

 surface of a soft pure metal instead of on a hard alloy such 

 as speculum metal, the form of the groove can be made to 

 conform more perfectly to that of the ruling point, which 

 moreover does not wear away rapidly as is the case when 

 the ruling is done on a hard alloy. In the case of the 

 echelette gratings I used a surface of gold-plated copper, 

 which gave such excellent results that I immediately tried 

 to find a surface suitable for optical gratings. A very thin 

 coating of silver was deposited upon a plate of polished 

 speculum metal by electrolysis, and was then very care- 

 fully polished. This was found to serve admirably, and a 

 trial grating, with 14,438 lines to the inch, was ruled with 

 the old Rowland machine, which showed a very dark central 

 image and a very bright first order spectrum. Spectra were 

 also obtained on plates prepared in this way which yielded 

 spectra in which certain colours were wholly absent, the 

 dark bands reminding one forcibly of the dark bands 

 observed with the curious grating described in 1902, except 

 that they were much wider. 



Dr. Anderson, who has had charge of the dividing engines, 



