THE 

 LONDON, EDINBURGH, and DUBLIN 



PHILOSOPHICAL MAGAZINE 



AND 



. JOUBNAL OF SCIENCE, 



SUPPLEMENT to VOL. XIII. FIETS SERIES. 



LXI. Preliminary Notice of the Results accomplished in the 

 Manufacture and Theory of Gratings for Optical purposes. 

 By Prof. H. A. Rowland, of the Johns Hopkins University , 

 Baltimore*. 



IT is not many years since physicists considered that a 

 spectroscope constructed of a large number of prisms 

 was the best and only instrument for viewing the spectrum 

 where great power was required. These instruments were 

 large and expensive, so that few physicists could possess them. 

 Professor Young was the first to discover that some of the 

 gratings of Mr. Rutherfurd showed more than any prism-spec- 

 troscope which had then been constructed. But all the gra- 

 tings which had been made up to that time were quite small, 

 say one inch square, whereas the power of a grating in resol- 

 ving the lines of the spectrum increases with the size. Mr. 

 Rutherfurd then attempted to make as large gratings as his 

 machine would allow, and produced some which were nearly 

 two inches square, though he was rarely successful above an 

 inch and three quarters, having about thirty thousand lines. 

 These gratings were on speculum-metal, and showed more of 

 the spectrum than had ever before been seen, and have, in 

 the hands of Young, Rutherfurd, Lockyer, and others, done 

 much good work for science. Many mechanics in this country 

 and in France and Grermany have sought to equal Mr. Ruther- 

 ford's gratings, but without success. 



Under these circumstances I have taken up the subject 

 with the resources at command in the physical laboratory of 

 the Johns Hopkins University. 



One of the problems to be solved in making a machine is 



to make a perfect screw; and this mechanics of all countries 



have sought to do for over a hundred years, and have failed. 



On thinking over the matter, I devised a plan whose details I 



* Communicated by the Author. 



Phil. Mag. 8. 5. No. 81. Suppl. Vol. 13. 2N 



