424 NATURAL SCIENCE [December 
selected types, but rather to show the relationships of animals and 
thus to manifest evolution in the Animal Kinedom. As the chief 
guide in determining those relationships, Prof. Roule takes em- 
bryology, although palaeontology and comparative morphology and 
histology are not rejected. 
The Animal Kingdom is divided into sixteen branches, dealt with 
in the following order: Sareodic Protozoa, Ciliate Protozoa, Mesozoa, 
Spongida, Hydrozoa, Seyphozoa, Plathelminthes, Nemathelminthes, 
Trochozoa (Rotifera; Bryozoa, Brachiopoda, Phoronidea, Sipunculidea ; 
Mollusca; Archiannelida, Hirudinea, Chaetopoda; Pseudannelida= 
Sternaspids & KEchiurians), Arthropoda, Chaetoenatha, Peripatida, 
Echinoderma, Enteropneusta, Tunicata, Vertebrata. Each of these 
branches is dealt with under the following heads: General considera- 
tions and relations to other branches ; Distribution in nature ; General 
organisation, first of the embryo, then of the adult ; Comparative 
account of the different body systems, as manifested in the various 
Classes of the Branch ; Principles of classification, division into Classes, 
and mutual relations of the Classes; Bibliography. The second 
volume ends with two indices, the first to the zoological names, the 
second to the anatomical terms. There is no general chapter dealing 
with the Animal Kingdom as a whole, or with the classification 
adopted, since that was given in the previous works referred to 
above. 
Seeing that the subject is one of such obscurity, and open to so 
wide diversity of opinion, it hardly seems worth while pointing out 
the paths along which we should not care to follow our professorial 
eulde. That he has made the attempt, and that Messrs Masson have 
published it, is alone a reason for gratitude. For the book, though 
somewhat wordy, and occasionally less clear than we are accustomed 
to from a Frenchman, furnishes a series of very readable accounts 
with many suggestions of interest. Without casting any slur ‘on 
embryological research, we must confess to some distrust in those 
who place quite as much reliance on it as does Prof. Roule. But taking 
it at his valuation, we fail to see how it lends support to the view 
that the Nautiloidea are ancestral to the Ammonoidea and the ad- 
mittedly dibranchiate forms ; for their embryology shows clearly that 
the Nautiloidea have lost an important structure, the protoconch, once 
possessed by them and still possessed by the other orders. We also 
venture to think that the known facts in the embryology of recent 
echinoderms afford no proof whatever that the five-rayed ancestor, 
which the holothurians must have had in common with the other 
classes, was less developed than many cystids. Prof. Roule’s ‘hypo- 
thetical Pentazodn’ is not the most ancestral form that is shadowed 
forth to us, either by embryology or by palaeontology. 
Special praise is due to the illustrations, which have nearly all 
been drawn for the work, under Prof. Roule’s direction, by Mr L. 
Jammes, in a style that is at once original and effective. We must, 
however, protest against the picture on p. 1275, purporting to repre- 
sent living crinoids, “dans un fond rocheux de convention” (penny 
peep-show convention). Attached to one of these marvellous rocks, 
by a stem far too short in proportion to-its arms, is a Pentacrinus ; 
how it is fixed one cannot tell, but certainly not by the cirri, as 
