: Ge 
== = @ n= zen = SE 
den ar den (1 —_ LS)" 
Further: 
EE Be 4 Be 
s_ — St = Saab, N Er = 
same 1 — wd)? wot Vo (rd)? 
or a term x>_ corresponding to every term #4, for which 
BI Sah Bi 7 and J es gee 
= 2 = Sp ES 
Ss eo NE 
(vr) ome 
EE 
Moreover 
1—1S=>1— Le = En 
n° +2 reed 
so that we find: 
2p(n* +2)? is 
= — = ii Sa byes = ENE SS Bede. 
Jen Dr) mm 
Zal 
If we call 4” the magnetic displacement for the unity of magnetic 
force, we find for the constant of rotation x: 
Of 20 O\2 
2p (n° +2)? | 1 — 
=a Sab = —¥ Bw) A), 
den (»,°— Pp")? ome 
en 
Representing. the sums 2 BA” by D we come to a formula: 
“ot 
2p (n° +2)? _ PD 
py td En Sa b ad 8 ais 
hd eg 2 2\2 
Jen (wv —v*) 
We shall now try to determine the constants C and D in such 
a way that these formulae represent the observations, the same free 
frequencies being taken for n and y. 
We shall then have to fix how many free frequencies we shall 
assume, as the number of terms in the sum 5, corresponds with 
this. We can be led in this by what is known for the dispersion 
of the index of refraction. As was already mentioned in the 
first communication, it is found that the dispersion of transparent 
substances can be explained by assuming a small number of ultra- 
red, and also one or more ultraviolet free frequencies. It must be 
tried, therefore, in the first place whether the assumption of one 
ultrared and one uliraviolet frequency leads to our purpose. 
Then two terms must oceur in the sum S for the index of 
refraction. In the usual way we shall suppose the red free frequency — 
so small that —»* may be substituted for »,*—v* for the frequencies 
60 
Proceedings Royal Acad. Aimslerdam. Vol. XVIII. 
