( 30 ) 
<p decreases from oo to 186. As, namely, here v remains nearly 
constantly = V* (0 remains practically =1), a l& will hardly change at 
all, so that the whole decrease of p is practically caused by the 
decrease of g>, i.e. by the increase of the small value of v—b [in 
consequence of the (slight) increase of v\. But in the neighbourhood 
of the minimum E y 0 will begin to change rapidly, and decrease 
from 1 to about 0 in the neighbourhood of the maximum D, on 
account of which the value of v will increase as rapidly from 7, to 
1, so that decreases from 10800 to 2700. Now in consequence 
of this rapid decrease of a ! v i, while <p changes little (at 7 7 =9from 
186 to 173) *), p will increase rapidly between E and D (see the 
Plate in the preceding paper). Between D and C we have again 
the same thing as between p = oo and the point E : 0 and v remain 
namely almost unchanged (resp. = 0 and 1), so that now for 
decreasing i p the value of °/»* changes again little, and p will only 
decrease through the decrease of y> (from 173 to 8); i.e. through the 
increase of the slight value of v — b in consequence of the (slight) 
increase of v. As we know, the increase of v—b will be less between 
Cand B than the decrease of a / v t in consequence of the great increase 
of v, so that p rises; while past B the further decrease of v — b or 
of v predominates, as then a / t > may be neglected by the side of —• 
-S-<— 
Further by logarithmical differentiation we find from ^ = A -: 
1—0 s (p 
5* 0(1 HP)/ 
4 2 [; 
IV) 
» that we find for 
which is always positive. 
of l—d has ? h!Ln Gf ' alS0 p 'i 766 ^ ° f the P recedin g paper the temporary deci 
££ ate^he“Lfn n, the rr The smal1 valoe of 1 
& and D, where both v and b rapidly increase — 
1 + * decreases in about the same degree, so that the fraction i±f=- 
L o^g deCreaSe ° f l ° WhiCh ^ 
