**The respiratory chamber open, viz. the front edge of the mantle is free from the back of the neck, leaving a large slit for the admission of the air into the bag. An operculum.

The families are 1. Cyclostomida; 2. Helicinidæ.

There can, I think, be only one opinion as to the great merits of this method, of which the author has given a most interesting exposition in the work referred to, one of small price and easy access. Its superiority as a whole to any previous one can scarcely be questioned; and the new views taken as to the position of several of the families in their respective orders, as well as in regard of the genera which are made to enter into the composition of the families, nothing less than the most extensive and critical knowledge of the entire class could have suggested. The method, Mr. Gray says, "is founded on the examination of the animals of all the molluscans contained in the London and Paris collections, as well as of all the drawings or engravings of the animals which I have been enabled to see, exceeding more than five thousand species, being at least one hundred times as many animals as were known when Lamarck proposed his system, and fifty times as many as were known to Cuvier when he published his system on the Animal Kingdom.”* The defects of it proceed from having had the attention too exclusively directed to the exterior anatomy of the animal irrespective of the form of the shell; whence it has resulted that the Haliotidæ are found alongside of the Trochusidæ, although their relationship is really distant, as Cuvier had proved. Other, and not less prominent, instances of misalliance might be pointed out, but the one selected has been made more apparent and decisive of late by the researches of Milne-Edwards and Emile Blanchard; and these researches, again, have effected a revolution in the arrangement of the class, which will, probably, receive the adhesion of future naturalists.

From the more or less perfect formation of the foot, which regulates the motions and much of the economy of the animal, Milne-Edwards proposes to divide the Gasteropoda into two subclasses, viz. 1. the normal Gasteropods, embracing the Pulmones, Nudibranches, Inferobranches, Tectibranches, Pectinibranches, Scutibranches, and the Cyclobranches of Cuvier; and 2. the aberrant Gasteropods or Heteropodes of the same author.

* Proceed. Zool. Soc. Lond. No. 178, p. 132.

The normal Gasteropods, although very numerous in species, are a natural group, exhibiting in its families, however, considerable variety in the organism, which appears even in the embryo at an early period of its development. In some the larva is furnished with a turbinated shell having the aperture closed by means of a little operculum; over the front of the head there is a large membranous veil more or less deeply divided into two lobes and garnished with a fringe of vibratile cilia, to make this veil an organ of locomotion;* and there is nothing to be observed that can be compared to an umbilical vesicle. In others the larva is naked; the head is not furnished with natatory veils with ciliated margins; and there exists, upon the anterior part of the dorsal region, a kind of umbilical vesicle.

The Gasteropods that, in these first steps of their development, affect these two forms, present also considerable anatomical and physiological differences when they have come to maturity. Some are pulmonated and breathe the unmingled air; others breathe water and are provided with gills. The first have been long separated from the branchiferous Gasteropods; but the close affinity which binds the latter together has not been sufficiently appreciated, nor indicated in our systems; for in all of them these mollusks are scattered throughout in a variable number of ordinal divisions, and no systematist has hitherto perceived that they constitute but one, although a large, group. Yet it is certain that, in their embryotic condition, the different families so much resemble each other that it would be difficult to distinguish generically the larvae of the Eolides, or of the Aplysiæ, from the larvae of the Buccina and of the Vermetus.

The normal branchiferous Gasteropods only differ from each other when they have advanced towards their ultimate forms; but the peculiar character of the heart, whose first existence is later here in life than in animals of higher organization, separates them into two natural groups, which ought, according to Milne-Edwards, to take the rank of Orders.

In one of these, named OPISTOBRANCHES, the blood is brought to the heart in a current directed more or less obliquely from behind forwards, and the auricle is usually placed in rear of the ventricle; the respiration is effectuated by the aid of arborescent or fasciculated branchiæ, which are not enclosed in an appropriated cavity, but are more or less exposed uncovered on the back or upon the sides, towards

*See Reid in Ann. and Mag. N. Hist. xvii. 382.

the hinder part of the body; the cervical region is always naked; the animals are hermaphrodite; and the shell, well developed on the larva, becomes rudimentary, or entirely disappears in the adult.

This order is constituted of Gasteropods, distributed in three of Cuvier's orders, viz. his Nudibranches, Inferobranches, and Tectibranches. In the classification of Lamarck we find them united together in the first section of Gasteropods; but they are mixed up with the Patellæ and the Chitons, whose structure is very alien. M. de Blainville again scattered them throughout four orders, which have no nigh affinity to each other; and some are widely sundered from their relations by the introduction of the Pteropods between the Aplysia and the Eolides. The Opistobranches form, however, a very natural group; and the characters which combine them together, as well as those which separate them from other Gasteropods, seem to justify their elevation to an ordinal rank in their class.

In the second division of the branchiferous Gasteropods, the abdominal portion of the body does not become rudimentary as in the Opistobranches, but is developed, on the contrary, in due proportion to the cephalic and pedal portions; and it is, throughout life, protected by a shell sufficiently large to allow the body to be drawn within its cavity. The cloak is always directed forwards, and forms, above the cervical region, a vaulted chamber of variable capacity where the excretory vents open externally, and in which the branchiæ are lodged. These respiratory organs are composed of simple lamellæ, laid parallel to each other, inserted along a vascular support, and affecting a pectinated or comblike figure. In general they are situated in front of the heart, and even when they are prolonged to the posterior part of the body, the branchio-cardiac vessels trend from before backwards, so that the blood comes to the heart in a current the reverse of that in the Opistobranches. Lastly, every species has its male and female individuals.

In the classification of Cuvier the Gasteropods that possess this assemblage of anatomical and physiological characters are disseminated in the orders Pectinibranches, Tubulibranches, Scutibranches, and Cyclobranches. Lamarck has arranged a part of them amongst his Gasteropods, and another part amongst his Trachelipods, where they mix but do not combine with the Pulmonata. M. de Blainville forms with them the first and the third of his subclasses of Paracephalophores, and intercalates between these two groups all the other Gasteropods. Milne-Edwards justly prefers to re

unite them in one and the same order, to which he gives the name of PROSOBRANCHES.

As to the Chitons, it is difficult to say where they can be placed most naturally in the system. Cuvier and Lamarck have concluded them to be affined to the Patellæ,-a decision in which concologists in general have acquiesced; but Blainville, being of opinion that they are not mollusks forms a class of them amongst annulose animals. To justify this view, he reminds us of the peculiar disposition of the valves of the Chitonidæ, which prompts a comparison with the segmented character of the Annelides; and of the position of the vent opposite to the mouth, which is likewise an annulose and not a molluscan structure. Milne-Edwards adds that the reproductive organs, while they differ essentially from those of Gasteropods, resemble those that are found in the Annelides. In the Gasteropods these organs are always unpaired and asymmetrical as well interiorly as in the position of their external orifices; but in the Chitons they are arranged alike on each side of the mesial line, with a couple of orifices similar to those of the Crustacea. The disposition of the circulating apparatus tends equally to alienate the Chitons from the Gasteropods, and to approximate them to the articulated animals, for the heart simulates a dorsal vessel, and has a structure very different from that of any normal Gasteropod. In fact, every thing in the organisation of the Chitonidæ appears to indicate a tendency to a bilateral distribution of the organs regulated by a straight line; while in the Gasteropoda the body as a whole, and in its parts, seems to have been modelled on a curved line. Thus the opinion of Blainville would appear to be more correct than that of Cuvier, as to the rank of the Chitons; but to solve the question it is first necessary to know the phases of these animals in their development: for it is their embryotic condition that can alone inform us if they are descendants from the Mollusca which have received some features of the annelidan race; or if they rather descend from the Annelides and ape only in part the habit of the Mollusca. Yet, however this question may be answered, it seems impossible to retain the Chitons any longer amongst normal Gasteropods: and were it decided to attach them to the Mollusca, it would be necessary to form with them an order apart, or rather to detach them as a small satellitious group, pendant from the typical body, without just being a constituent portion of it. Mac Leay would call it an osculant group; but Milne-Edwards, more fanciful than he, would liken it to a satellite that occupies a thinlystudded field in space, removed far away from the closely

stellar constellations that crowd the glorious vault of heaven. -Guided by these views, Milne-Edwards classifies the Gasteropods thus:-

Class-GASTEROPODA.

Cephalous Mollusca, with a fleshy foot for locomotion, formed by a posterior lobe of the head; the organs of generation unpaired and unsymmetrical, and the entire organism modelled on a spiral line, either in the larva state only or throughout life.

Typical Group, or Subclass I. NORMAL GASTEROPODS.

Foot fleshy, flattened, and very large; abdomen well developed, &c.

1 Section. PULMONATED GASTEROPODS.

Larva with a naked head, &c.; vessels of the lesser circulation reticulated; androgynous.

2 Section. BRANCHIFEROUS GASTEROPODS.

Larva with cephalic swimmers, &c.; vessels of the lesser circulation fasciculated.

Order I. OPISTOBRANCHIA. Cervical region naked, &c. Order II. PROSOBRANCHIA. Cervical region surmounted

with a vaulted pallial cavity, &c.

Aberrant Group, or Sub-class II.-SWIMMING GASTEROPODS, OR HETEROPODS.

Foot fleshy, vertical; abdomen rudimentary, &c.

Satellitious group of GASTEROPODS pendulous from the
Prosobranchia.

Family, CHITONIDE. Cephalous Mollusca ? with a fleshy foot; the body subannular; generative organs paired and symmetrical: a medial dorsal vessel, &c.*

These great reforms of Milne-Edwards derive support from the delicate and elaborate dissections of his young friend, M. Emile Blanchard. This adroit and gifted naturalist, in addition to the characters above assigned to the Opistobranches, tells us that the cephalic ganglia of the nervous system have a closer centralisation than they have in other Gasteropods, a fact which had been previously made known to us, in many Nudibranches, by Mr. Hancock and Dr. Emble

* Annales des Sciences Naturelles, ix. (1848), p. 102-112. It appears the classification was first published in August, 1846-Mr. J. E. Gray has defined the genera of Chitonidae in the Ann. and Mag. N. Hist. xx. 66.

Q Q