84. KANSAS UNIVERSITY QUARTERLY. 
In making his estimates of visibility the observer was kept in 
ignorance of the field strength or whether the magnet was excited 
at all. The visibilities for the second set of observations, which 
were taken on another day, are all multiplied by $6 before plotting 
in order to reduce them to the same initial ordinate as the others. 
RESULTS. 
In Fig. 2 curves a to e are the visibility curves for maxima, plot- 
ted from the preceding observations. 
Curve a, Magnetic Field 0. GG Siunits, 
Curve by Magnetic Bield 1166 °: * 6 
Curve c, Magnetic Field ino py A Mlle i “ 
Curve d, Magnetic Field 2950 © ‘‘ es 
Curve e, Magnetic Field 7843 ‘‘ aL 
It is seen that the visibility curves fall off more and more rapidly 
for increasing field strengths, indicating a general broadening of 
the spectral line. Curve a agrees very closely with the dotted 
D? 
; 4 We , avinee 
curve a’, whose equation is V==.76 «2 and from which is cal- 
culated the half width of one of the sodium lines to be 6,==. 249° 
0123 wave numbers, which can be reduced to Angstrém units by 
multiplying by the factor 3.47 which gives 6,==.0425 Angstrom 
units. Hence the distribution of light in a single sodium line 
free from the effect of magnetism is given by the expression 
x2 
Peas: 8. which is plotted as curve a Fig. 3. 
yo |. Bay Oa Mena 
ue 
6 ja o idl . 
%. 
DM sy iy Balive 
a 
4 —<— Ne oie 
ig | 
ten 1 
2 — -_——} _— _ anemone 
IN * 
"50096 " FIV. Fi GOp 
Ain Angstrom unrs 
