Mea.<in'('i)ie)its of Small I^esistances. 33 



between them, and the intensity of the smalle.st of these 

 maxima will be I-. 



The dependence of the width of the lines in the difFraction- 

 speetrnm on the elfective width of the grating is easily shown 

 as follows : — Suppose it is assumed that the first maximum 

 practically tades ont of sight when n((^ + <^^) difi*ers from 



37J. 



27rnpj its value at the principal maxima, by + -y , in which 



case the intensity is less than one-tenth that of the maximum 



Then, 



n7r(a + h) sin . , 37r 



-^-^ = i-rnp±^, 



and . ^ 2\p 3X 



de= "^'L^l = 3\ ^ 3X 



COS ^ 2n{a-j-b) cos 6 2iv cos 0' 



where ic represents the width of the grating. 



The angular width of a line in the spectrum is 2d9 or 



3X, 



IV COS 



assuming that the line fades out where the intensity is one- 

 tenth of the maximum. 



In ruled gratings, whether transmitting or reflecting, the 

 value of / depends on the nature of the grooves cut by the 

 ruling diamond as well as on the angle 6. and therefore the 

 expression given above for /, which assumes a simj^le slit^ 

 does not apply. But the summation holds good in any case 

 if the ruling is uniform. 

 Amherst College, U.S.A. 



IV. Measurements of Small Resistances. 

 By Albert Campbell, B.A.'^ 



THE object of this paper is to give a brief account of a 

 number of measurements of a set of low-resistance 

 standards belonging to the National Physical Laboratory ; 

 these tests were made partly with a view to comparing various 

 methods of measurement, and it is from this point of view 

 that they derive their main interest. The resistances were of 

 the oil-cooled type and had separate potential terminals. 

 They were made of manganin, and their nominal values were 



* Communicated by the Physical Society : lead March 13, 1903. 

 PhU, Mag. S. 6. Vol. 6. No. 31. July 11)03. D 



