JUNE 13, 1884.] 
praise than that he has bestowed upon the 
German naturalist. 
The plan of the work we are reviewing is 
excellent. In place of attempting a hard and 
fast definition of the molluscan phylum, he has 
described and figured an architypal mollusk in 
detail; and the reader, once familiarized with 
this type, can follow clearly the discussion of 
the subordinate branches. These are taken up 
seriatim, beginning with the more archaic forms. 
The phylum is divided by Lankester into two 
great branches,— the Glossophora and the Lipo- 
cephala. The first comprises three classes, — 
Gastropoda (in its widest sense) ; Scaphopoda, 
or the Dentalia ; and Cephalopoda, with which 
the author includes. not only the cuttles, etc., 
but the Pteropoda. The Lipocephala are 
equivalent to the Acephala of Cuvier, or the 
Lamellibranchiata of authors. 
So far, the position of the Pteropoda (though 
not absolutely new) with the Cephalopoda, 
rather than as a class by themselves or as a 
subclass of Gastropoda, is the chief difference 
from the generally accepted classifications, and 
one which will be much criticised, if not finally 
rejected ; since the little that is known of the 
embryology of the pteropods differs in impor- 
tant features from that of the cephalopods. 
The Gastropoda are first discussed in a 
general way, and, on the whole, in a most satis- 
factory manner. We could wish, that, in intro- 
ducing new Greek derivatives, some attention 
had been paid to euphony ; for surely that me- 
lodious language can afford better terms than 
gonad (‘ sexual organ’), osphradiwm (‘ sensory 
organ’), ctenidium (‘gill’), and others which 
grate upontheear. We note among details the 
erroneous statement that the radula is horny (it 
is really chitinous), and that the jaws are usu- 
ally calcified, and almost universally present. 
No single instance of a calcified jaw among 
recent Mollusca occurs to us, and there are 
large groups without a jaw. ‘The jaw is com- 
posed of a substance allied to chitine, more or 
less combined with really horny material, the 
former defying alike the strongest acids and 
alkalies to reduce it. 
The recognition of the radula as a feature of 
the highest systematic importance is very wel- 
come : it is only to be regretted that the author 
seems to have fallen into utter confusion in his 
endeavors to indicate formulae for the teeth, 
and to have followed, without much investi- 
gation, the crude notions of Dr. Macdonald, 
rather than the researches of Troschel, Lovén, 
Woodward, or Sars. It seems also to have 
escaped him, that the radula is occasionally 
(though rarely) absent.* 
SCIENCE. 
731 
The author divides the Gastropoda into two 
subclasses, Isopleura and Anisopleura, charac- 
terized by the relations of the organs, which, in 
the former and more archaic group, are bilater- 
ally symmetrical with a posterior anus: in the 
Anisopleura the visceral mass has been subjected 
to torsion, bringing the anus to the anterior 
right side, while the concomitant twisting of 
the remainder of the intestinal mass results in 
a masking of the original symmetry. In the 
process when originally brought about, if the 
termination of the intestine was sufficiently 
low, it became entangled in the pedal nervous 
loop, which, in following it, acquired a figure-of- 
eight form. If, on the other hand, the plane 
of intestinal torsion was above the pedal loop, 
the latter did not participate in the torsion, and 
in the succeeding generations retained its simple 
character. These relative features, observed 
by Spengel in a number of mollusks, are made, 
after him, the occasion of two super-ordinal 
groups, — Streptoneura and Euthyneura. 
We regard the establishment of the sub- 
classes above mentioned as a decided advance 
on previous systems, while it is doubtful if the 
super-orders will stand the test of future inves- 
tigation. The character adopted as a basis is 
purely mechanical, and, so far as yet shown, 
without serious significance. 
The isopleurous gastropods comprise the 
Chitons, Neomeniidae, and Chaetoderma, which 
are considered respectively as typifying orders. 
In our opinion, they should have been divided 
into two super-orders, — one the Polyplacoph- 
ora, exhibiting a metameric repetition of the 
primitive shell-sac, and possessing a developed 
and functional foot; the other the Chaetoder- 
mida, without (as adults) a primitive shell-sac 
or shell of any kind, and with the foot aborted, 
or rudimental. The statement(p. 641) that the 
cuticular spines of the latter group ‘ replace’ the 
shell, is not correct in a strictly scientific sense, 
and the expression were better not used; for 
these spines are absolutely identical with the 
spines of the girdle in Chiton, and have no re- 
lation to the true shell. 
The Anisopleura Streptoneura are divided 
into two orders, Zygo- and Azygo-branchia, 
accordingly as the suppression of the origi- 
nally left-side organs is or is not carried out. 
These characters we regard as unsatisfactory, 
and the division resulting as artificial; Halio- 
tis, Fissurella, etc., being combined with the 
true limpets, while their (to our notion) much 
nearer relatives, the Trochidae, Pleurotomari- 
idae, etc., are left in the other order. Recent 
observations on the limpets indicate that this 
arrangément cannot be maintained, though we 
