108 ON A TABLE OF STANDARD WAVE LENGTHS OF THE SPECTRAL LINES. 
Wt. 
Angstrom .... 5895.81 1 
Muller and Kempf . 5896.25 2 
Kurlbaum .... 5895.90 2 
Peirce 5896.20 .5 
Bell 5896.20 10 
Mean, 5896.156 in air at 20° and 760 mm. pressure. 
Tills must be very nearly right, find I believe the wave length to be as well 
(letenninecl as tlic length of most standard bars. Indeed, farther discussion of the 
question would involve a very elaborate discussion of standard meters, a question 
involving endless dispute. • I think we may say that the above result is within 1 in 
100,000 of the correct value, which is very nearly the lirnit of accuracy of linear 
measurements. This should be so, as the probable error of the ano-ular 
o-^ me 
nects the wave length only to 1 in 2,000,000,* and hence nearly the whole accuracy 
re?«ts on the 
meas 
Kovl.t 
RESUME OF Process for obtaining Eelatiye Wave Lengths. 
1 
1. Determination of about 20 lines in the visible spectrum by coincidences by 
o 
I 
Determination of about 15 lines in the visible spectrum by coincidences by 
Rowhuiu, usmg several gratings of 21^ feet focus 
3. Interpolation by direct eye observations with concave gratings of 211 feet 
focus and micrometer of 5 inches range and of almost perfect accuracy. 
4. I 
by means of ilat ij-ratino-s. 
J3...^...g 
5. Mea«,retnout of pliotographic plates from 10 to 19 inches 1 
or tljree 
having 
of the spectrum in different orders on them, thus connecting the 
Ultra rmlet and bh,e witli the visible spectrum. The fact that nearly the same 
values are obtamod for the violet and ultra violet by use of different parts of the 
visible spectrun, proves the accuracy of the latter, 
6. Measurement of pbotographic plates having the solar 
ible spectr 
mcdence wUh the metal lines of different orders of spectra. The fact that the 
temperaLre? V^Js ^^^^^^^'P''*'!^™^*^^ «uis the best standard of length, and almost independent of the 
- -- ••" - . caUbrarion of cZu.Tf.r ?'" "T ^^ '^^'^ °" ^"^^ "^^ ^"S^^^ ^ith almost perfect accuracy, 
practical use. 
t Tliese 
nspd. They serve a „..f i ? "^ """^^^^ S^ven no weight, on 
bey serve a useful purpose, however, as checks on the other ;ork. 
LIS 
h- 
