( 800 ) 
di; d 
= —(. For as was proved in 5, 5 = 0 lies here at smaller volumes 
UV U 
dp dl : dp . a : 
than = — 0, and so in the region where = is positive. The line 
da lax 
dp Re 8 E 
Er — 0 gives the minimum value of the pressure for every definite 
HH 
x, and so this minimum value becomes constantly higher towards the 
right, or in other words an isobar of a certain negative value of p 
cannot penetrate further to the right than that x, for which this 
dp; aen OD 
value is reached on — =0. At the intersection with — = 0 the 
dv da 
isobar is // the z-axis, and afterwards it reaches the line z = a,. 
The course outside this quadrant has no physical sign;ficance. So we 
get a course as fig. 10 presents for an isobar very close to «= 7, 
Fig. 10. 
But not all the isobars originating from & =@,, v =O will present 
this course. The question, however, what the shape of the others is, 
and where the limit lies between the different kinds, can only be 
discussed when we have gained a complete survey of the points of 
d d, 
intersection of = and — 0, also of those lying at a great 
& Vv 
distance. 
10. The results obtained up to now are pretty well independent 
of the question whether 6 is a linear, or a quadratic function of 
x, but now we must distinguish between these two suppositions. 
For it is clear that for points lying considerably to the right many 
