640 
NATURE 
[AUGUST 12, 1915 

rectilinear movement; but apart from this, the 
use of the word “rectilinear” here is a petitio 
principii of the grossest description. How can 
we say anything about the adaptation of the 
bones to rectilinear movement until we have the 
concept of what rectilinear motion is? 
This leads us to the consideration of the more 
metaphysical parts of the book. Frankly, we 
disagree with the author on one or two funda- 
mental points. He protests against the absolute 
as an illusion—even against what we may call 
the special absolute; we believe, on the other hand, 
with Bradley, that in a certain sense the absolute 
does “precede,” or is involved by, the relative. 
At the same time, we may fully admit that 
concepts are formed by a psychological process, 
and that there is no realisation of an absolute, in 
the same sort of way that we cannot realise the 
arithmetical continuum as we can the numbers 
2and 3. Again, the author just touches on cases 
of illusion, insanity, and complete hallucinations ; 
saying of the latter that “fortunately they seem 
quite rare.” Doubtless this is fortunate, in one 
sense; but to ignore them without attempting to 
give a philosophical explanation of them is merely 
a confession of failure. However, we must agree 
to differ on metaphysics for the presént, and be 
thankful that so much has been done recently to 
clear away fallacies and quibbles, and bring the real 
problems of philosophy into view. . : 
The translation reads well on the whole, and we 
have only found one misprint of importance: on 
p. 321 “vertical” should be “‘vortical.” On p. 17 
the account of Russell’s “vicious circle’ analysis 
is not clear; whether this is the fault of the author 
or that of the translator we cannot say. What 
Russell has pointed out is that we are talking 
nonsense when we ascribe to a class of classes a 
defining quality which characterises its members : 
for instance, to speak of “the number of all 
numbers” is as nonsensical as to say “the animal 
comprising all animals.” Similarly, there is no 
“class of all classes,” and this is the most 
important example of the theorem. G. B. M. 

PHYSICAL CHEMISTRY OF MOLECULES. 
Molecular Association. By Dr. W. E. S. Turner. 
Pp. vili+170. (London: Longmans, Green and 
Cos; 1915:)) Wrices5s-enet. 
E progress of chemical theory has been con- 
nected in an intimate way with the study of 
the complexity of molecules. By the introduction 
of Avogadro’s hypothesis the determination of 
molecular complexity in gases was reduced to a 
comparatively simple problem. The development 
of the osmotic theory of solutions provided a 
NO. 2389, VOL. 95] 

theoretical basis for the interpretation of the re- 
sults obtained in the investigation of the vapour 
pressures and freezing points of solutions. In 
so far as dilute solutions are concerned, the pro- 
blem of molecular complexity is partially solved 
by the application of osmotic methods, although 
the question of the association of solute and 
solvent in the same molecule does not lend itself to 
attack in this way. When we pass from dilute to 
concentrated solutions or liquid mixtures, the 
problem acquires an entirely different character. 
For such mixtures there is no general guiding and 
reconciling principle such as is afforded by the 
hypothesis of Avogadro. The same difficulty 
confronts us when we deal with pure liquids. The 
methods available for the investigation of the 
molecular condition of pure liquid substances are 
entirely empirical, and the value to be attached 
to the various methods which have been proposed 
is at present largely a matter of personal opinion. 
It is with the dependence of the complexity of 
the molecule on the nature of the constituent 
atoms, and on the forces which act on it, in so far 
as these forces are modified by changes in tem- 
perature, concentration, state of aggregation, and 
nature of the solvent, that the author’s subject- 
matter is concerned. One chapter is devoted to 
the molecular complexity of gases, two to the com- 
plexity of dissolved substances, and three to that 
of pure liquids. 
The molecular condition of pure liquids has 
attracted a great deal of attention in recent years, 
and the literature of the subject has attained to 
such dimensions that a summary and critical sur- 
vey of the methods involved is particularly oppor- 
tune. Molecular association has been the theme 
of the author’s own work for many years, and it 
is therefore quite intelligible that the monograph 
is not a mere compilation of facts and hypo- 
theses, but represents an extensive, well-ordered, 
and closely reasoned discussion of the salient 
features of the subject. 
In an appendix, which runs to forty-four pages, 
the author gives a tabular record of the results of 
the investigation of the molecular complexity of 
dissolved substances. As a table of reference this 
should be found extremely useful, and its com- 
pilation adds materially to the value of the work. 
Comprehensive as the monograph appears to be, 
there are certain noteworthy omissions. For in- 
stance, the reviewer has looked in vain for any 
reference to the determination of the molecular 
complexity of dissolved substances by means of 
the lowering of transition temperatures. Such 
investigations afford information relative to the 
molecular condition of substances in concentrated. 
salt solutions. Although limited in their appli- 

