( 138 ) 
d(AE) 
f0F dT~ ~ ^ ® which has the value — oo for 7=0, 
as (? is always positive); then reaches a maximum negative value 
(evidently when log T = — ^tf), increases again for higher 
values of T, becomes again =0 (when log T= — , and then 
remains always positive, and that increasing, because the increase of 
T and that of ^( — A£) w ni exC eed the decrease of 
p? 0' in consequence of the continual decrease of the pressure of 
coexistence p. 
In the expression 
dp _ A E 
dT 7\—A V) 
LE 
— will now be of the order «' -j- £ log T in the neighbourhood of 
T= 0, so that then —~ is also of the order «' -f- ^log T, because 
— AF remains finite. In other words: — is = + od for 7’=0; 
dT ^ 
becomes = 0 for log T = -, where AT? becomes = 0 for the 
ft 
second time (see above), and will then become negative and continue 
to decrease, because past the minimum of A E ^for log 
A E continues to increase (see above), while we have already seen 
that A V is a quantity decreasing with T. 
So this course of ~ gives for the line SM (coexistence liquid- 
solid) for negatieve values of A b and AF a course as represented 
in Fig. 6 of the plate. (S is the triple-point). Hence the line SMN 
will touch the pressure axis in the point N (7 T =0), because there 
J ) Disregarding the logarithmical order of A E, it has been erroneously derived 
in II, p. 36 [formule (12a) and (13)] that the limiting value of A E for 0 
would be yRT , and therefore that of would have been given by - 
