1391 
2. the bivariant regions 135, 123 and 134 occupy all spaces, so 
that they form an angle of 360° around the invariant point. 
3. the occurring curves and regions belong to one another in 
such a way that the eurves are the limits of the regions and there 
occur as many curves as regions. 
These properties are not only valid for the example, discussed 
here, but in general we may express them in the following way. 
1. A complex XY (represented by a point Y of the concentration- 
diagram) can be converted into several monovariant equilibria 
M, M,... and into as’ many bivariant equilibria B, B,...; it 
depends on the P and 7, which we give to the complex X which 
of these equilibria occurs; under each P and 7’ only one of those 
equilibria occurs. 
2. In the P,7-diagram the bivariant-regions B, B,... do not 
coincide neither completely nor partly; nor do they cover one 
of the curves J/, M,...; these curves separate the regions from 
one another. ; 
3. The bivariant-regions 5, B,... occupy together all spaces of 
the P,7-diagram, so that they form an angle of 360° round the 
Invariant point. 
We shall call those properties the rules of the fields. 
In order to show this we draw in the P,7-diagram a closed curve, 
which surrounds the invariant point; this closed curve intersects 
all fields and curves of the diagram, therefore also the fields 
B, B,... and the curves M, M,... When we change the Pand 7’ 
of a complex Y in such a way that it follows the closed curve, then 
consequently at this cycle all equilibria M, M,... and B, B,... 
must arise from the complex \ in a definite order of succession. 
As, of course, in this case no other equilibria than those mentioned 
above, may be formed, the fields 5, B,... have to cover this closed 
curve completely, so that the fields form a continuous figure round 
the invariant point. Also it appears that the conversion of a biva- 
riant equilibrium into the next one does not take place directly, but 
first a monovariant equilibrium is formed from the bivariant one 
and from this the other bivariant equilibrium. The properties 
mentioned above, follow at once from those considerations. 
In the different diagrams we see the confirmation of the rules 
mentioned above. Let us take e.g. a binary system [fig. 2 (I)]. 
A point N on the line A B, between the points /, and F, can 
represent one of the monovariant equilibria (/’,) = /, F, F,.(/,) = 
= FF, F,F, and (F, =F, F, Ff, or one of the bivariant equilibria 
