eee 
_ W. Gibbs on a norm ap of te ay Oa, i ee 
strém after undergoing a correction which depends upon the © 
value of the wave length of the more refrangible lineof D. This 
value is taken by Ditscheiner as 588°8, the wave length as deter- _ 
mined by Fraunhofer. But as, for the reasons above mentioned z 
I consider the results of Angstrém more reliable, I have taken 
his value, 589°43, in place of that of Fraunhofer, and have r 
duced all the measurements of Ditscheiner in accordance wit 
this assumption. The following table will suffice to show thi 
agreement between the two series of measurements. 
Tasie [. 
Angstrém, m | Ditscheiner. | Dir. Angstrdm, Ditscheiner. Diff. 
687°49 687-81 0732 529-32 Td +o°10 
656°76 656°66 —O8) 523-69 523-74 +005 
619'17 61972 | 40°55 51963 519°67 +0°04 
617°08 617750 +0°42 518 79 518-73 +0°06 
616°33 616°73 +0°40 517°66 5770 +0°04 
61431 614:24 =0'O7 Sra 45 517°15 fore) 
612°34 612°7 +0°42 492°23 492°22 -O-0O1 
610°45 610°82 +0°36 4gi-At 4gi'4r 
5g0°04 590°07 +003 48g°50 48954 +0°04 
589:43 589° 43 487 55 487755 00 
561 56 +0'0r 486752 486°49 -o0:03 
560-70 560-77 +0'07 44o'81 440°87 +0°06. 
559:10 559°13 +003 438-63 438°75 | +012 2 : 
557-66 557°77 +or11 434:28 4:34 06 
545-97 546 05 +0:08 432°78 432°82 "O04 
545-07 545-08 +001 431-03 431°35 +0°32 
534-41 534:41 000 39771 7°10. - 
533-16 533-29 | +013 393-59 39374 | +0715 
528-73 528-79 +o 06 
A number of Ditscheiner’s measurements refer to lines which 
cannot certainly be identified with the lines ae by partic- 
ular elements. These lie between G and f have been omit-_ 
ted as useless for our present purpose. But in the greater 
iber of cases the wave lengths in his ae of results are re- | 
aha directly to the lines given in Kirchhoff’s chart, which not 
rely enables us to identify them, but pernaits us to obtain the i 
ware Beceihs of the intermediate lines by interpolation. 
following table gives the corrected values of all the wave lengths _ 
measured by Ditscheiner with the exception of those between G 
and H, and ‘of those of the nine principal lines of Fraunhofe 
which are given in oe I* For these I have taken the valu 
given by Angst I have not ventured to make a further use 
of Angstrém’s deotiuiekioes for the reason that it is onl 
th rough a comparison with the measurements of Ditscheiner that 
Gn 
they can be identified with the lines upon Kirchhoff’s seale. In 
every part of the spectrum excepting that between C and D the 
agreement between the two sets of measurements appears to fall 
within the limits of the errors of observation. 
* Marked in Table II with asterisks, 
