742 BELL SYSTEM TECHNICAL JOURNAL 



chosen for work upon the longer X-rays, as optical gratings are ruled 

 with lines unusually far apart for work in the near infra-red. 



The ruling of good gratings is an art ; and those who have practiced 

 it with conspicuous success are fewer far than those who have 

 attained pre-eminence in music or in painting. Amateurs, mechanics, 

 and professors figure upon the list, the first of all being Fraunhofer, 

 who from a glazier's apprentice evolved into the founder of spectros- 

 copy. After his gratings of wires and of scratches in a foil of gold-leaf, 

 he invented the method of engraving with a diamond-point upon a 

 surface of metal or of glass (he used the latter) which is followed to 

 this day. He met and grappled with all the difficulties which were 

 later to beset his followers, and described them in language which now 

 sounds strangely modern. Then, as now, it was possible to rule tens 

 of thousands of rough grooves roughly to the inch; the trouble lay, 

 as still it lies, in making them identical and spacing them equally. 



Equality of spacing depends upon a screw, which is turned through 

 a prearranged angle and is expected to advance through a definite 

 distance carrying the future grating with it, whenever the diamond has 

 completed one ruling and is waiting to begin the next. Screws as 

 manufactured are not good enough; and anyone who aspires to be a 

 maker of gratings must first of all procure the best available, and then 

 devote a long and tedious time — literally years — to making it still 

 better. Primacy in the art passed to America in the eighties of the 

 last century, because Rowland of Johns Hopkins developed with 

 much labour a process for removing, or at least for mitigating, the 

 imperfections of a screw. The greater the number of rulings to be 

 laid down side by side, the longer the portion of the screw which must 

 be made, as nearly as humanly possible, perfect; and Michelson has 

 testified, from unrivalled experience of many years, that the time 

 required for the process varies as the cube of the length of the screw 

 and width of the planned-for grating. Increase of resolving-power 

 thus is bought at an enormous price in patience and in perseverance. 

 A research institute is as proud of a notable grating by Rowland or 

 Michelson or Wood, as a picture gallery of an authentic Titian or 

 Velasquez; and the promise of a new talent is not more joyfully 

 received, than a rumour that someone is working to perfect a yet longer 

 screw to make a yet wider grating. 



Alikeness of successive rulings depends on the endurance of the 

 diamond. The ruling-engine is sequestered in a well-insulated room, 

 and after the temperature has settled down to constancy is set in 

 motion by some device worked from outside, and left to do its task 

 in solitude. If the diamond breaks, or suffers any great change in 



