A WAVE-LENGTH COMPARATOR FOR STANDARDS OF LENGTH. 31 



on the scale of 40,000 to' the inch, satisfactory from every point of view, not only for 

 the purpose in question of an end-defining line on a standard bar, but also for an 

 ulterior purpose which the author had in view ; this was the engraving on a standard 

 bar itself of approximately wave-length rulings at stated intervals, which could be 

 used instead of the glass etalons of MICHELSON and of FABRY and PEROT, in a simpler 

 and more direct method of determining the total number of wave-lengths of a 

 particular radiation, red cadmium light in all probability as the most homogeneous 

 yet known in the Standard Yard. The surface of the speculum metal was finely 

 polished as truly plane as possible by Mr. GRAYSON himself before the rulings were 

 engraved by his own uniquely delicate hands, with the aid of his wonderful ruling 

 machine ; and for the sake of certainty of preservation of the surfaces they were 

 covered with thin cover glasses. 



The results with all the other rulings, of three different finenesses, and on glass and 

 the various metals mentioned, were also more or less satisfactory, particularly the 

 40,000-to-the-inch rulings. Hence it may be safely taken as quite possible to engrave 

 five rulings on the scale of 40,000 to the inch on glass and on several of the metals 

 employed for standards of length. The rulings on speculum metal, however, are so 

 superior that the author has decided to make detailed experiments with them, and 

 Mr. GRAYSON has made for the Standards Department a considerable number of such 

 rulings on 20 different small but relatively thick plates of speculum metal, covered as 

 described, and of two types. One type, of which there are 15 similar specimens, consists 

 of a plate 9 mm. square and 2 mm. thick at the base, but bevelled along the edges 

 for insertion in a dovetail in mounting on the bar. Each lias ruled upon it a single 

 "location signal" as represented in fig. 13, the five parallel lines being on the scale of 

 40,000 to the inch. The other type, of which there are 5 specimens, is a plate ot 

 equal width and thickness to those just mentioned, but 34 mm. long. On each of 

 these there are ruled location signals at the following intervals : A zero point, 

 3\ inch, Yj inch, ^ inch, ^ inch, |- inch, and 1 inch, and also an additional signal at 

 ^ inch on the other side of the zero location signal. The reason for this is that the 

 total interval of 1 inch and ^ is an aliquot part of a yard, which may be reached by 

 a repeated process of doubling ; for its double is 2-, the double of this is 4^, twice 

 this is 9, which again doubled gives 18, and finally the double of this makes up the 

 36 inches of the yard. 



These rulings are infinitely superior to those of NOBERT. The author is particularly 

 struck with the accuracy with which Mr. GRAYSON who has entered into the author's 

 ideas most warmly and put his very highest skill at the disposal of the Board ot 

 Trade has attained the exact spacing of his rulings ; for the author has rigorously 

 tested them by direct measurement against interference bands. It is interesting that 

 lines 40,000 to the inch are almost precisely wave-lengths of red light. The exact 

 wave-length of red C-hydrogen light is, in British measure, ^r?To mcn > an< ^ tnat ^ 

 the red ray of cadmium is 3-9 jy^ inch. There was a time, not very long ago, when it 



