931 
squares. We found 
en EE 
n° +2 ‘ 47,485 . 10%? _— p? y? 
NS a an ge 
nt 2 A8 161, 10°° D? py? 
The agreement appears from the foregoing table which shows 
that a representation of the two quantities with the same free 
frequencies is possible. 
it may be further examined in how far the found numerical 
values of the coefficients are in agreement with what theory would 
lead us to expect. For a comparison with the theory it is necessary 
to recalculate the constant of rotation in the new theoretical unities 
used by Lorentz. We find then: 
2p? (n° +2)? hf 312,37.10°° 
Dr 0080 ee 
Yen (47,485.109°—p?)? 
2p? (n° +2) 307,84.10°° 
(11) ee a) 0,05556 — —_ | 19-25, 
Jen (48,16 1.10 —p°)? 
A comparison with the above derived theoretical formulae yields 
first of all (the indices #24 being further omitted) : 
XE B, Aa = 0,06702.10—25 
1 
(I) - .— > By = 0,28204, 
Va vat 
(11) — 0,27587, — 0,05556.10—25 
or if we introduce a mean value A,,, for the different values Aq: 
ORS SO. Ase S Be = 00040210 inf 
(11) = 2. 0,27587 ra’, = 0,05556.10—75 vof 
from which follows 
Aam = U) 1,584.10—26 va’, am = (IL) 1,348.10-2 vq?. 
If we take for A,, the normal value of the magnetic resolution 
Aam = 3,19.107, 
we find: 
(1) Va = 20141022; DB ride 
(1) ono 02 — 9830 
and from this the wavelength: 
ha = (Z)0,0000042 em. = (ZZ) 0,0000039 em. 
which really represents a wavelength very far in the ultraviolet. 
The following terms give 
(J) 2 > By = 2,4850.10°°, = Bj Ay = — 312,37 10° 
(11) — 2,6182.10®, — 307,84.10°° 
| 
