442 Mr. M. Ponton on the Law of the Wave-lengths 
These differences are so far within the limits of probable errors 
of observation, being all of them less than the least of the differ- 
ences between the corresponding members of the two observed 
series, that there need be no hesitation in admitting them for 
the sake of obtaining a series so regular as the foregoing, and 
presenting the peculiar advantage of rendering the whole of the 
waye-lengths deducible from that of either B or D. 
In all calculations involving these wave-lengths, it will be 
found more conyenient to adopt, instead of the actual lengths 
corresponding to any standard of mensuration, the re/ative wave- 
lengths referred to that of B as unity, stating the others in frac- 
tional parts; thus keeping the numbers independent of any 
standard of linear measure. The wave-lengths and their loga- 
rithms will then stand as under. 
Relative wave-lengths referred to B as unity. 
Logarithms. Numbers. 
C 1:9794999 0:9538934 
D 1:9325036 0:8560588 
E 1°8834154 0°7645667 
F 1:8477024 0°7042103 
G 1:7947653 06233979 
H 1:°7563732 05706545 
Mean wave M 1:970]116 0:9334940 
It remains to compare the values of the wave-lengths corre- 
sponding to the fixed lines, with those found by Newton’s series 
for the boundary lines of the coloured spaces of the solar spec- 
trum; and for this purpose the latter must be reduced to the 
standard of the French inch, when they will be found to stand 
as under :— 
Fixed lines mean ob. 
B 0:000025410 
Newton’s wave-lengths. 
0:000024929 
eh }oonoo2so C 0-000024235 
walla Mire we D 0:000021750 
Green. . { 1900020579 E 0:000019440 
Bie os lenis F 0-000017915 
Indigo .  poneeabbe 
Violet. . { 0:000016987 G 0:000015860 
0-000015705 H 0:000014575 
