156 On the Aherrat'wn of the Concave Grating. 



criticize these views very stronprly, for the reason that, as 

 pointed out at the beginning of the paper, I appear to be 

 the first to have pii])lisheJ the general theory of the instru- 

 ment and to have advocated its use. I can only say that 

 since then mv own views have ])een decidedly modified by a 

 further experimental and mathematical investigation of the 

 capabilities o£ the instrument. After the publication of Poor 

 raid Mitchell's first paper, I pointed out some of the ob- 

 jections to the instrument as used by them in photographing 

 stellar spectra, on the score of its low light efficiency as 

 compared with other forms of stellar spectrographs, par- 

 ticularly the transmission (wire) objective grating *. It was 

 not until later, in the use of the instrument in solar eclipse 

 work, that its limitations in the directions of resolving-power 

 and photographic field were clearly recognized. This led me, 

 .ns already stated, to the present investigation. It now^ appears 

 from this that perhaps the most serious objection of all to the 

 instrument is in its lack of accuracy. This point has been 

 very briefly touched upon in this paper and in the one on 

 ^' Conditions of Accuracy, &c.'^ t already referred to, but will 

 be more fully developed in a paper dealing w^ith a discussion 

 of Frost's and Humphrey's results on the flash-spectra %. 



In conclusion we may quote with a considerable degree of 

 pertinency to the present question the remark of Schuster 

 with reference to the prism-spectroscope. He says ('Enc. 

 Brit/ vol. xxii. p. 373) : — " The dispersion and magnifying- 

 power of a prism can be considerably altered by a change in 

 its position, and a knowdedge of this fact is of great value to 

 an experienced observer. The use of a prism in a position 

 difi'erent from that of minimum de^dation is, however, a 

 luxury which only those acguainted icith the laws of optics can 

 indulge in with safety.'^ Change '^ prism " to grating, and 

 " minimum deviation " to Rowland's mounting, and the 

 applicability of Schuster's statement to the case we have been 

 discussing is very evident. 



Alleglienv Observatorv. 

 1901-1902. 



* See " Note on tlie Use of the Grating in Stellar Spectroscopic Work," 

 Astroplivsical Journal, yoI. vii. p. 198; also "The Objective Spectro- 

 scope," ibid. vol. It. p. 75 et seq. 



t AstropliTsical Jonrnal, vol. xvi. pp. 286-288. On p. 288 there is a 

 typographical error which is corrected in the March number of the 

 Jom-ual. The values of ^mai. as given in the December number are ten 

 times too lars-e. i. e., in place of 2^ 20' and 6^ the values of ^max should 

 be0=-22and0°-6. 



X " On Measurements of Wave-Length with the Concave Grating 

 Objective Spectroscope," Astrophysical Journal, vol. xviii. p. 1, July 1903. 



