TRANSACTIONS OF SECTION B. 505 
The author suggests another explanation, which involves the following suppc- 
sitions :— 
(a) Below the critical temperature the interfacial tension between the two 
phases is positive for all values of the radius of curvature. 
(6) At the critical temperature the interfacial tension becomes zero for all 
ordinary curvatures, but remains positive for very small values of the radius of 
curvature. It will be seen that this assumption involves the furthér one, that at 
the temperature of disappearance of the meniscus at a bounding surface of 
ordinary curvature (critical temperature) the two phases do not become identical. 
(c) At temperatures slightly above the critical the interfacial tension is still 
positive for very small radii of curvature, but negative for all ordinary curvatures. 
(d) At still higher temperatures the interfacial tension becomes negative for 
all curvatures, 
These assumptions carry with them the main assumption that the interfacial 
tension between two liquid phases increases in general with diminution of the 
radius of curvature of the bounding surface. This increase, however, need not 
extend to the very smallest values of the radius of curvature. It is also neces- 
sary to suppose that the curve connecting interfacial tension with radius of curva- 
ture of the bounding interface suffers a shift towards the region of negative values 
of the interfacial tension as the temperature rises, 
Granting these assumptions, the existence of permanent opalescent states above 
the ordinary critical point follows at once. For it can be shown that the con- 
ditions specified, for example, under (c) would produce a system in which one 
phase would be distributed throughout the other in a state of very fine sub- 
division. If the particles are small enough, such a system would not present a 
milky appearance, but would doubtless be opalescent. 
An interesting question concerning the nature of the critical state is thus 
raised—namely, as to whether there is not sufficient experimental evidence to 
justify the belief that the passage of systems through critical states is insufficiently 
described by the simple theory of Andrews? That is the wider question at issue. 
The particular form of explanation suggested in the present paper also raises the 
question as to whether the observed phenomena may not admit of interpretation 
by taking into consideration the operation of capillary forces ? 
These points seem worthy of discussion, especially in relation to the theories 
of ‘ gaseous’ and ‘ liquid’ molecules—‘ Gasonen ’ and ‘ Fluidonen’ devised in recent 
years by de Heen, Traube, and others. 
The present communication has arisen out of a correspondence on the subject 
with Professor van’t Hoff, to whom the main idea is due. 
FRIDAY, AUGUST 19. 
The following Papers and Report were read :— 
1. On Crystal Structure and its Relation to Chemical Constitution, 
By Professor Paut Grotu. 
The molecular hypothesis assumes that in solid bodies the molecular move- 
ments occur about certain equilibrium positions which can only be altered by the 
operation of external forces. In crystalline bodies the spacial arrangement of 
these equilibrium positions must consequently be a regular one. 
The possible kinds of crystal structure—that is, the kinds of regular arrange- 
ment cf congruent molecules—were first discussed by Bravais, who described 
fourteen such kinds of structure and distinguished them as ‘ space lattices.’ 
Bravais’s theory, however, only explains seven kinds of symmetry which possess 
the properties of crystal symmetry, and the incompleteness of his work reposes on 
the introduction of an unnecessary assumption, namely, that the molecules occupy 
parallel positions. Sohncke discarded this assumption in attacking the problem, 
