Standards of Optical Wave-length. d71 
thousandths A,a great advance will have been made, and 
many questions that are now wunapproachable can be 
answered. 
Of course all the measurements of spectra already made— 
and I have myself worked a good deal in that direction— 
will be useless as far as wave-lengths are concerned. Prof. 
Hartmann, of Potsdam, thinks it already now possible to 
caleulate fa Perot al Fabry’s measurements a table of 
corrections, by the application of which Rowland’s measure- 
ments and all the others based on Rowland’s standards may 
be corrected. I think that is impossible. We do not even 
know if the fluctuations in Rowland’s system are truly 
known, as very few lines have been remeasured by Fabry 
and Perot ; and they may have fallen on some lines with 
large individual errors, and the curve of corrections may 
become quite another one by the use of other lines. But 
even if that were not the case, who knows from what in- 
dividual standards every line of a spectrum has been deter- 
mined? Hasselberg is the only worker who gives in his tables 
the standards between which he has interpolated, so that his 
numbers can be corrected when corrections to Rowland’s 
standards are known. ‘The greater errors of individual lines 
in Rowland’s iables cannot be corrected without remeasuring 
all the lines. And, further, it is quite impossible to deduce 
any table of corrections for the ultra-violet part. As the 
method of coincidences has broken down for gratings which 
have been tested, we can reach no conclusion as to the 
behaviour of Rowland’s system in this part of the spectrum 
without new absolute determinations. 
So I think that Rowland’s system must be entirely aban- 
doned; and a new system must be created. As for all spec- 
troscopic purposes only relatively correct values are needed, the 
new system might be based on Rowland’s false value for the 
D lines; then new measurements would coincide within some 
hundredths A with the old ones. I think it more logical, if 
an entirely new system is to be created, to base it on 
Michelson’s absolutely exact values of the Cd lines. But 
this question is of course of minor importance; of great 
importance is only, that the new system should be inter- 
national. 
