purely vegetable diet; but their true molars are differently formed, and the second even departs from this type of dentition, by having only two incisors in the upper as well as in the under jaw.

I

This form of dentition is common to the genera Phalangista, Petaurus, Phascolarctos, Macropus, and Hypsiprymnus. The next organs to be considered as influencing the habits and economy of the marsupials, are the extremities. I have already observed how wofully these organs have been neglected by the makers of systems; though it is difficult to conceive how scientific zoologists could possibly undervalue or overlook the instruments of the most important and striking actions of animal life. Not only do the great function of locomotion, and its thousand varieties and adaptations, whether to aquatic, arboreal, terrestrial, or aerial habits, depend solely and entirely upon the extremities; but the scarcely less important functions of prehension, manipulation, burrowing, and even the sense of touch, the source of our most excellent ideas, and the index of intellectual power, reside in the same organs. The formation and modifications of the extremities, therefore, do not furnish those merely second-rate characters, which should justify the philosophical zoologist in postponing them to slight modifications of dentition, or neglecting them altogether. On the contrary, as every action and habit of animal life, except the mere appetite, depend upon these organs; as they are the most extensively influential, so their modifications should hold the most prominent place in every system, and will be invariably found to lead to the most natural and philosophical arrangements.

The marsupial quadrupeds, always excluding the Monotremata, which cannot be properly compared with other mammals, present four very distinct and primary influential modifications of the extremities.

1. The Pedimanous form, where the fingers are long, separate and prehensile, and the hind thumb opposable to the other toes. The animals consequently possess perfect powers of prehension and manipulation; they are entirely arboreal, feed indifferently upon vegetable and animal substances, though preferring the former, and all have a cæcum of moderate dimensions. This family, which I shall denominate Cheirogrades, from their locomotion being performed by

I use this term as synonymous with the common word, handling. In the paper as originally written, these families were called respectively Scansores, Cursores, Saltatores, and Fossores: I have now substituted the names in the text, as more pliant when used adjectively, as in speaking of saltigrade or digitigrade marsupials, &c. April, 1839.

means of hands, comprises the genera Didelphis and Cheironectes, with didelphoid teeth; and Phalangista, Petaurus, and Phascolarctos, with macropoid.

2. The digitigrade form, in which the functions of prehension and manipulation are very much impaired, or altogether absent. The hind feet are without opposable thumbs, which, however, are sometimes represented by a small, motionless tubercle; the animals tread only on the toes in walking, and their pace is confined to the surface of the earth.— All are characterised by the didelphoid system of dentition, a regimen principally confined to animal substances, [and an entire absence of cæcum]. This family, which, from the nature of their pace, I shall denominate Digitigrades, comprehends the genera Thylacinus, Dasyurus, Phascogale and Myrmecobius. 3

I

2

3. The saltigrade form, in which the posterior extremities so immeasurably exceed the anterior in length, as to preclude the ordinary mode of progression on all fours, and to compel the animals to proceed by a series of successive springs, sometimes from the long hind legs only, sometimes from the hind to the fore legs, as in the hares and rabbits. The toes of the fore feet are separate and prehensile, and the animals enjoy perfect powers of manipulation; but the conformation of the hind toes is altogether unique among mammals. The thumb is tuberculous, or altogether wanting; the two following toes are small, slender and inclosed in the same skin, being marked externally only by their double claw; the ring finger is the largest of all, of a size altogether disproportioned to the other toes, and armed with a powerful triangular claw; and the last, or outer finger is of intermediate size, and provided with a similar claw to that just described. The dentition comprehends examples of both the systems above characterised; the food of some genera is consequently mixed, though in all cases it is principally composed of vegetable substances, and a cæcum, sometimes of very large dimensions and complicated form, is invariably present. This family, which I shall call Saltigrades, contains the genera Macropus and Hypsiprymnus with macropoid teeth, and Perameles and Charopus with didelphoid.

2

4. The plantigrade form, which is confined to the single genus Phascolomys. The toes here are short, rigid and unprehensile, well adapted for burrowing, and without any power of manipulation. The animal treads on the entire sole of the

1 I have inserted this fact on the authority of Dr. Grant.

3 and These two genera have been discovered since the paper was written, and are now inserted for the first time.-W. O. April, 1839.

foot in walking, the pace is slow and confined to the surface of the earth, and the toes are so firmly united as to be altogether destitute of separate motion. The dentition, as far as regards the number of incisors, may be called rodent, though it is really very different from that of the true Rodentia; the food is exclusively vegetable, and the alimentary canal is characterised by the presence of a capacious and complicated cæcum. I distinguish this family by the name of Plantigrades.

The following table exhibits these relations in a more condensed form; and will give a good idea of what I consider to be the most natural and philosophical arrangement of the marsupials. Except the kangaroos, they are all of nocturnal habits.

Having thus traced the progressive history of the general distribution of marsupial quadrupeds, inter se, it remains for me to offer a few observations upon the rank which the main group itself ought to occupy among the natural families of the animal kingdom. M. De Blainville, in his view of the subject, seems almost disposed to regard the marsupials as forming a distinct class, parallel to and co-ordinate with the Mammalia themselves; and in this sentiment he has been, to a certain extent, followed by Baron Cuvier. Illiger, less happy in fixing their position in the graduated scale of existence, than in defining their generic differences, distributes them throughout three different orders; and Latreille, whilst he regards the marsupials generally as forming a distinct natural order, considers the Monotremata as a separate CLASS,

intermediate, in rank and position, between birds and mammals. Finally, Baron Cuvier, in the second edition of the Règne Animal,' adopts the most judicious part of this arrangement, by separating the common marsupials from his extensive order Carnassiers, with which he had formerly associated them, to elevate them to the rank of a separate order: still, however, retaining the Monotremata as a family of the order Edentata. In this view he had been already preceded by Temminck, excepting that the eminent zoologist considers both the Monotremata and the ordinary marsupials as distinct orders, equivalent to other groups of the same rank and denomination.

This arrangement appears to me to be more consistent with the order which nature has herself established, than any other which has been yet proposed; unless that I am disposed, after the example of M. De Blainville, to unite the Monotremata with the other marsupials, rather than to continue them as a subordinate group among the Edentata. In fact, so long as the possession of mammary glands is considered as the distinctive and peculiar characteristic of the class of mammals, so long should the singular modification of these organs and of their functions, exhibited in the marsupials, entitle those animals to rank as a primary division, or order of mammals: but I can in no case consider them as an equal and coordinate group, or CLASS, since their distinctive characteristic is but a subordinate modification of the general type of organic structure, common to all mammiferous quadrupeds.

With regard to the Monotremata, also, though the question of their viviparous or oviparous production still remains undecided, I can, under no circumstances, regard them as a parallel and equivalent group to mammals, birds, and reptiles. Meckel distinctly asserts the existence of mammary glands in the female Ornithorhynchus ; and this circumstance alone, even though the mamma exist merely in a rudimentary form, and without the accompaniment of the ordinary function, I esteem sufficient to determine the rank of the Monotremata as a subordinate group of mammals.2 In fact the simple de

I

1 The observations of Meckel have been fully and most satisfactorily confirmed, since this passage was written, by the investigations of Mr. Owen; and it is now definitely established that these singular and anomalous animals, not only lay eggs and hatch them like birds, but likewise support their young, when excluded from the shell, by means of a thick milky fluid, which at that period exudes copiously from the glands observed by these able anatomists.

2 For the following ingenious observations on this subject I am indebted to the kindness of Professor Agardh, now Bishop of Bergen: and it affords VOL. III.-No. 31. N. s.

2 N

finition of this class, as mammals, or animals provided with

me great satisfaction to find that my views regarding the value and position of the group, Monotremata, so entirely coincide with those of so distinguished a naturalist. They are contained in a letter dated Lund, in Sweden, Aug. 3rd, 1833; and are a translation from a work recently published by Professor Agardh, in the Swedish language, under the name of 'Allman Wext Biologi.'

"The marsupials," says the Professor, "are Mammalia which approach very nearly to birds; the Monotremata in particular almost coincide with them. Not only do the developed form of the hind legs, the deranged functions of the anterior extremities, the position of the body, and the destination of the tail to govern the pace, all indicate this affinity, but their internal structure is likewise very similar. They constitute a distinct group of Mammalia, combining carnivorous as well as herbivorous animals, in the same manner as birds contain predacious as well as frugivorous tribes. They have no distinct internal uterus, for it is only the connection of the two oviducts to which that name has hitherto been given; neither have they a peculiar vagina, for the organ which Daubenton and Geoffroy thus distinguish, when they assert that the marsupials have two vagina, belongs rather, according to the researches of Tyson, to the oviducts or Fallopian tubes: so that, except in the doubleness of the parts, the marsupials resemble birds in their organs of generation, as well as in other respects. The embryo also is brought forth, not as in other Mammalia, perfectly formed; but it is produced in the state of an egg, and in that form deposited in the marsupium or uterus. Now the egg or embryo of the Mammalia has the property of attaching itself to every part of the uterus at the point where the placenta is formed; and thus the embryo or egg of the marsupials fastens itself to the mamma, and there communicates with the arteria epigastrica in the same manner as in other Mammalia it communicates with the arteria uterina. It is fastened by a cord resembling the navel-string, (though it is unknown where this cord passes out from the embryo), which is often so long that the embryo hangs out of the bag, and which at the moment of real birth is separated by a rupture, as in the case of the placenta and ordinary uterus. This external uterus, however, does not invariably assume the form of a purse or bag; in some instances it consists of simple folds of the skin, and in the monotremes, even these disappear.

"The monotremes bear a very strong affinity to the ordinary marsupials. they likewise very closely resemble birds, not alone in the construction of the bill, cranium, clavicles, shoulder-bones, sternum, and undeveloped teats, but especially in their organs of generation. These animals have only one ovarium developed, as in birds, and both the Echidna and the Ornithorhynchus lay eggs and hatch them. Thus it is that the uterus of the Mammalia becomes modified in the marsupials, so as to be situated without the body, and finally vanishes altogether in the Monotremata.

"If we apply these considerations to ascertain the concatenation of the various groups of animals, in relation to their organs of generation, we find that it indicates one class, the Mammalia, which have an internal hatching organ, called the uterus; another class, the marsupials and monotremes, in which this hatching organ is placed without the body, vanishing totally in the latter group, the animals of which lay eggs and hatch them; and finally, a third class, birds, in which this property, which is irregular and limited in the monotremes, becomes fully normal."-Allman Wext Biologi af C. A. Agardh, p. 453.