604 Messrs. Aichi and Tanakadate: Theory of 
expected. Also for the smaller value of K, 7. e. of 7, the 
same reasoning will hold true. _ 
From this approximate expression, 
: sin am (=") Sin {,/3nK®\/ K6} 
Take eR 
it follows that F does not increase at maxima to 2X (mean 
: i 
term), but only to (mean term + 5~- ; and at 
dT@KDKE 
minima it does not diminish to zero, but only to 
1 
mean term — 5aKoOKe)} moreover, for values of 5 ne 
which sin {,/37K®,/K6}<0, the maxima of sin 27 & y 
changes to minima and minima to maxima. Finally, the 
expression for the intensity being 
73 
(0) =const. (=) f?(K@) . . for point source, 
I(6, ®) =const. Ga) F(K, ®, K@) for circular source ; 
it follows, firstly, that for larger value of ®, the difference of 
1(@) and I(@, ®) becomes larger ; secondly, that for larger 
value of 7, the difference of maximum and minimum values_ 
of i(@) becomes larger in virtue of 75, but for 1(0, ®) at the 
same time it is diminished by the presence of F. 
6. We shall now treat the case which has often been tested 
by experiment with the glass rod, ana straight slit as the 
source of light. If we take into account the breadth of the 
slit, there is no difficulty in applying the similar reasoning as 
above, to arrive at the expression 
r' \3 “t 
1(@, &) =const. (Gx) F(K, ®, K@), 
+K® 
F(K,: ®, p)— f? (KE + 2)dz, 
—K® 
where 2@=the breadth of the slit as viewed at the place of 
the glass rod. 
In § 5 we always substituted the mean value of \/(K¢)*—= 
before integration, so that the expression for F becomes only 
roughly approximate; but in the present case, there being 
no such term as »/(K¢)?—2z?, this expression for F must be 
