148 
INGA OTE 
| DECEMBER 13, 1906 
to that of a saline electrolyte—that is, it reacts rather 
as a solution-aggregate than by a succession of 
molecular combinations; the masses actually reacting 
following the stoichiometrical ratios proper to the 
dimensions of these ultimate groups, and retaining 
their relationship in the aggregate, which is thus 
progressively modified by the entrance of the new 
groups ”’ (p. 7). 
Owing to the prevailing ignorance as to the nature 
of colloids and the relation of this condition of a 
substance to its chemical character, both the language 
and the ideas employed by the authors in the develop- 
ment of their thesis are, as they themselves admit, 
somewhat vague, and it is difficult to realise exactly 
wherein lies the advantage of the new standpoint over 
the old view of cellulose as a highly complex mole- 
cule, coupled with the recognition of the fact that 
both the parent substance and many of its derivatives 
are only known as colloids. There can, however, be 
no doubt that sufficient attention has not hitherto 
been paid to this cardinal fact of the colloidal 
character of cellulose, and the authors do good 
service by insisting upon it and showing very clearly 
how this conception may serve to suggest many 
hopeful lines of investigation on questions of scientific 
and technical importance. 
The first section of the book contains: the develop- 
ment of these ideas, together with a general account 
of the chemistry of cellulose. In the second section 
are brought together the more important researches 
on the subject of cellulose which have appeared during 
the period 1900-5. An impartial abstract of each in- 
vestigation is given, followed by critical notes on the 
bearing of the results on the great question of the 
chemical structure of cellulose. The third and con- 
cluding section deals with the progress made on the 
technical side of the subject during the same period. 
This book therefore forms a supplement to the two 
volumes which have preceded it, but it is valuable, 
not merely as a compendium of the latest researches 
on cellulose, but much more as a thoughtful and 
suggestive contribution to our knowledge of the 
chemical and physical structure of this important 
natural product. ArTHUR HARDEN. 
OUR BOOK SHELF. 
Cours d’Astronomie. Premiére partie: Astronomie 
Theorique. By H. Andoyer. Pp. 221. (Paris: A. 
Hermann, 1906.) Price 9 francs. 
Tuere is no preface to explain the scope of Prof. 
Andoyer’s book, but it appears to consist of the notes 
of a course of lectures on spherical astronomy. Now, 
it is characteristic of lecture-notes to offer definitions 
in place of explanations; also, they have a tendency to 
disintegrate into a bewildering array of unconnected 
problems. The book has these defects. But as an 
exposition of the art of manipulating the very cum- 
bersome formulz of spherical trigonometry which per- 
vade astronomy, it will fulfil a useful purpose. The 
mathematical treatment is good and concise; more- 
over, the problems treated are mostly of a severely 
practical character. The author has wisely taken as 
his. guide the Connaissance des Temps; he refers to 
NO: 1937, VOL: 75] 
it continually, and there is very little in the book 
which has not some direct bearing on the use or con- 
struction of its tables. : 
The usual subjects are fully treated, refraction, 
parallax, aberration, precession, and nutation; there 
is a brief account of motion in an ellipse. The chap- 
ter on the geocentric motions of the planets is not 
very satisfactory; the student who has followed the 
lengthy investigations of the preceding chapters might 
safely have been offered something more advanced 
and more approximate to the practical problem than 
the very rudimentary theory here given. The ap- 
parent motion of satellites is in like manner inade- 
quately treated. The last chapter, which deals with 
eclipses, is, perhaps, the best feature of the book; 
solar eclipses are treated in a very thorough and 
interesting way. The general accuracy and precision 
of the book are admirable; the approximations and 
assumptions made are always clearly stated. Occa 
sionally, however, precision is carried to excess, as 
for instance, when the proper motion of Arcturus is 
given in seconds per tropical year (p. 141). 
It is a pity that the book is not printed in the usual 
way. It appears to have been reproduced in facsimile 
from the written manuscript. This is a needless sacri- 
fice of clearness, and must to some extent diminish its 
value as a book for reference. ‘ 
Les Révélations de l’Ecriture d’aprés un Contréle scien- 
tifique. By Alfred Binet. Pp. viiit+260. (Paris: 
Félix Alcan, 1906.) Price 5 francs. 
In this book M. Binet, the well-known experimental 
psychologist of the Sorbonne, describes an investigation 
of the art of telling intelligence and character from 
handwriting. After some preliminary inquiries to 
ascertain how far ‘‘ graphologists ’’ are able to recog- 
nise sex and age by means of writing, M. Binet sub- 
mitted to several experts specimens of the handwriting 
of people of great intellectual eminence, such as Renan, 
Dumas fils, and Claude Bernard, together with others 
obtained from persons known to be of ordinary intel- 
ligence. The general result was to show that, though 
the experts were more often right than wrong, they 
were liable to the grossest errors, as in one case in 
which Renan was judged to be of mediocre and un- 
cultivated intelligence, an opinion into which the ex- 
pert appears to have been led by the repetition of a 
word in the sample. 
In the estimation of character a similar result was 
obtained. This was tested by submitting to the 
graphologists specimens of the handwriting of 
notorious criminals to be distinguished from the writing 
of people of good moral reputation. Here again the 
experts were usually more or less right, but bad mis- 
tales were made, as when a man who had murdered 
his mother with every circumstance of brutality was 
judged to be a young girl, ‘‘ douce, modeste, et peu 
coquette.’’ In the simple distinction of the two classes 
of people concerned in the tests, the number of correct 
answers was distinctly greater than should have been 
expected from chance, and this preponderance of cor- 
rect judgments was greater in the estimation of in- 
telligence than in that of character; but it is not clear 
that clues derived from the subject-matter of the 
samples of writing were altogether excluded in the 
former case. 
In M. Binet’s hands the graphologists themselves 
became the subjects of investigation, and it may per- 
haps be regarded as evidence that their art has a scien- 
tific basis that some of the experts showed themselves 
greatly superior to others, under an experimental pro- 
cedure which deprived them of many of those adventi- 
tious aids on which it is probable they usually rely. 
