so 



and 



( 77 ) 



X 



Rb,h {^) + Sb,h (.i-) = ^' ii\k) = Qij, (.t-), 



Sb,h (a-) _ I ^ Qb,h {ic) _ Pbji ('^OA 



X 2\x ^ J 



(23) 

 (24) 



As now according to (17) (after substitution of ?i = 0) 



lim — '- = 



a:=oo X 



and on the other hand according to (21) and (22) when the limit 

 is put equal to X^.a, 



lim — = Lb^h"^ 0, 



we find from (23) and (24) 



km = - Lb^h , 



,. Sb,h{x) 1 



km ^ — =z - Lb,h , 



so 



,. Pb,h{''«) T 



WOT = 1 . 



a=oo Sb^h (•'^•) 



This confirms the supposition expressed by Mr. Kluyver at the 

 conclusion of his paper: 



If we divide the integers mb -f- h without quadratic factors into 

 two classes according to their consisting of an even or an odd 

 number of prime factors, the ratio of the numbers of integers minor 

 to X of both classes converges for .^=qo towards the limit unity. 



Berlin, May 23'^ 1904. 



Crystallography. — "On Benzylphtalimide and Benzylphtal-iso'' 

 imide" By Dr. F. M. Jaeger. (Communicated by Professor 

 A. P. N. Franchimont. 



(Communicated in tlie meeting of May 28, 1904). 



Some time ago a preparation was forwarded to me by Prof. Dr. 



Gabriel of Berlin, of a Benzylphtalimide'^): Q>J1^.{C0)^.^{Q,R^.0^M^) 



melting point 115°, 5, which he had obtained by synthesis. A sliort 



time afterwards, tiirough the kindness of Prof. Hoogewerff, at Delft, 



1) Bed. Ber. 20. 2227. 



