ON THE ASTIGMATISM OF ROWLAND'S CONCAVE GRATINGS. 5 



Any incident ray passing through the diaphragm over (under) the 

 line CQ and through the slit will come at a focus in the lower 

 (upper) half of the field. A short but rather broad prism, 2 or 3 

 mm. in height, placed at Q and reflecting lateral solar light, will 

 give a narrow solar spectrum with perfectly defined edges, passing 

 through the centre of the field; at the same time it will obstruct 

 none of such rays, emanating from a sodium-flame or arc-light placed 

 somewhere about T, as may concur in forming a sodium- or metal- 

 spectrum in the remaining part of the field. Of course if we wish 

 to get the metal-spectrum as bright as possible, the cone of light 

 furnished by the condensing lens S must be wide enough to fill up 

 the wedge formed by the rectangular diaphragm and the slit. 



With the third and ulterior spectra and with a very large grating 

 the condensing lens should be of rather great dimensions, so I think 

 the method will only be quite applicable with the first and second 

 spectra. I may add that probably the very best plan would be to 

 have a bicylindrical lens, or two cylindrical lenses put crosswise, of 

 such a curvature that both its orthogonal caustics might coincide 

 with the above named caustics of the grating; but every different 

 angle v or at least every successive spectrum would require its 

 especial lens. 



Through the kind permission and efficacious assistance of Professor 

 Hag a I have been able to control the above by a provisional exper- 

 iment. A narrow central band of the field on a black ground 

 showed the first sodium-spectrum originating from a strip of mirror- 

 glass, height 2,5 mm., placed along the caustic at Q, at 171 mm. 

 from the slit, and upon which the light of a lateral Bunsen-flame 

 was concentrated through a lens, f = 150 mm. The strip of glass 

 just arrested the superfluous central part of a direct beam of sun- 

 light that filled out the upper and lower parts of the field with its 

 spectrum. The sunlight had to be passed through several layers of 

 wire-gauze in order to bring down its intensity to that of the reflect- 

 ed sodium-light. Now in the compound spectrum the two positive 

 sodium-lines ended abruptly where the negative sodium-lines began ; 

 yet two very narrow sharp black lines, about 0,1 mm. wide, sepa- 

 rated the three contiguous spectral bands: this was occasioned by 

 the strip of glass having been simply cut with a diamond without 

 any ulterior grinding or polishing; so the somewhat rugged edges, 

 while they were unable to take part in the reflection of the sodium - 

 flame only acted as a barrier against the sunlight grazing them. 



In order to try to what limit, if need be, the method can be 

 applied, we turned the moveable girder of the spectroscope on to 



