L.Bell — Absolute Wavelength of Light. 281 



It is possible, however, to check this result by referring S a 2 

 to the Berlin platinum standard through the medium of the 

 Coast Survey meter " No. 49." This latter standard was com- 

 pared in 1876 with meter 1605 and directly with the platinum 

 meter. The details are given in Prof. Foerster's report con- 

 tained in the report of the Coast Survey for 1876. The result 

 of the direct comparison was 



PI— «49"=+24-4>. 



But now Prof. Rogers has compared R 2 with " 49," obtain- 

 ing in terms of the assumed length of R 2 



" 49 " = A— 19-3//, the assumed value of R 2 

 was, R 2 =A + l - 3yU. Hence we have, 



R„ — "49" = 20-6//, from which follows, 

 PI— R 2 = 3\S//. If now the equation 



between PI and the Metre des Archives established by direct 

 comparison in 1860 be correct : 



A — Pl = — 3-01/f. And therefore, 



R 2 — A = — 0-8/^, a result which is in 



close accordance with those derived from the Conservatoire 

 meter and Type I of the International Bureau by means of 

 the Standards T. and C. S. 



In my final determination of wave-length, I have used the 

 mean value of S a 2 as derived by the foregoing methods. Col- 

 lecting equations, 



S% + 0-96a<=| A 



From T. 



S% + l-04/.<=|A 



" C. S. 



S%+.l-40/*=fA 



" " 49.' 



S*-l-68 ={A„ 



" R . 



Giving to the equations derived from C. S. and R 78 twice the 

 weight of the others, we have finally, 



S% + 0-27/./=|A . 



I have given the relations derived from C. S. and R 78 double 

 weight because these standards have been compared directly 

 with the standard of the International Bureau, which now, 

 probably, should be regarded- as the ultimate standard of 

 reference. Especially is this true, since it is rumored, appar- 

 ently not without foundation, that the Metre des Archives is, 

 at present, for some unassigned reason, undesirable as a direct 

 standard of reference. 



It is unfortunate that there is not more general uniformity 

 in the material, shape and mass of standards of length. Dif- 

 ference in these particulars are fruitful sources of error in com- 

 parisons, and when one adds to this the trouble arising from 



Am. Jour. Sci.— Third Series, Vol. XXXV, No. 208.— April, 1888. 

 17 



