The Elephant might have been expected to have had a single superior cava, as have some of the largest existing Rodents, the South American Subungulata, and all other existing large-sized mammals. But it has, like all the Old World and all Nearctic Rodents, two. And it appears that there is some reason for holding that the Proboscidea, in contradistinction to the living Perissodactyla, and to the Artiodactylous Camel, Pig, and Deer, are an Old World type, and to be expected therefore to follow that type in such particulars as the one specified.

The Elephant however has the symphysis of the lower jaw perfectly anchy-losed; it is more entirely testicondous; its brain is richly convoluted, as indeed are the brains of most existing mammals of large size except the Manatee, and its placenta is zonary.

The zonary form of the placenta similarly distinguishes the Hyracoidea from the Rodentia whilst uniting them more or less with the Proboscidea. In the structural composition of the malar arch Hyrax agrees with the two orders just mentioned, and differs from the Ungulata, but on an estimate such as is given by Brandt (Mem. Acad. Sci. St. Petersbourg, 1869, Tom. xiv) of the sum total of the affinities of this animal, it should be ranked as an Ungulate with Rodent ward affinities rather than as a Rodent. Brandt's own words run thus, p. 119: 'Es werden daher die Hyracen im Allgemeinen wohl am passendsten als Ungulata gliriformia oder glirioidea bezeichnet und als eine, jedoch weit mehr den Huft-thieren ahnliche, Mittelform zwischen den genannten Thieren und den Glires angesehen werden konnen.' On another page, p. 116, Brandt suggests that Hyrax may connect the 'Pachyderms,' by which he in this connection means the Perissodactyla, with the Sciuromorphous Rodents specially, and also with the Lagomor-phous, whilst the Mesotherium would stand similarly in relation to the Lagomorphi and the larger animals in question.

Andreas Wagner indeed had expressed himself in 1844 in opposition to Cuvier, and to the same effect as Brandt, by saying that a separate family should be created for Hyrax amongst the Pachyderms, and that it should be considered as forming a transition towards the Rodents.

Cuvier, as is well known, in opposition to the view which trivial names, such as that of iMarmotte batard; given to Hyrax, embodied, went so far as to speak of it as being a dwarf Rhinoceros, and in his Ossemens Fossiles, ii. p. 127 seqq. 1822, he enumerates certain points of resemblance between the two animals. These points are such as the number, 20-21, of the ribs, as the transverse direction of the condyle of the lower jaw, as the absence of canines, as the shape of the nails, and as the presence of but three toes on the hind feet, and are inadequate to the support of such a view as the epigrammatically stated one just quoted. His suggestion as to placing Hyrax between Rhinoceros and Tapir is less open to objection. It may be remarked that some of the peculiarities of Hyrax, such as the bisacculate character of its stomach, the presence of a sterno-maxillary muscle, the dilatations of its Eustachian tubes, the flatness of the roof of its skull, the perfect orbital ring, the absence of a clavicle, and of an acromion on the scapula, and the presence of a diastema between the incisors and the canines, are points which are at least as indicative of equine affinities as of connection either with the Rhinoceros or with the Rodent stock.

And if on the other hand the presence of a languette on the dorsal, and of papillae foliatae on the lateral surface of the tongue in Hyrax, are points curiously reproduced in the Rabbit and the Rat, the important point of the absence of a second superior vena cava distinguishes Hyrax from all rodents except the Caviidae, and the zonular character of its placenta distinguishes it essentially from all known Rodents whatever.

Few points of real affinity connect the Rodentia with the Insectivora in addition to those more superficial peculiarities in general appearance, size, and habits, which have caused the two orders to be connected in common language. And these points are mainly such as may be characterized as indications of comparatively low organization in the scale of Mammalian life; and they are rarely constant in all the members of the two orders. Among them we may specify the not infrequent vacuolation or fenestration of the bony roof of the palate, the imperfect condition of the bony support of the tympanum, and the retention of the primitive jugular foramen. Similarly in the soft parts of both orders we find usually, if not always, cerebral hemispheres devoid of convolutions, and two superior cavae, as we do in all known Sauropsida. But irrespective of the differences in dentition, which may appear to lose some of their importance since we have been acquainted with the existence of Apatemys from the Middle Eocene of America, which had (see Marsh, l. c. p. 43) gliriform incisors combined with insectivorous molars, the digestive organs in the two orders are strikingly different, the intestinal tract being provided with a caecum in all Rodents except the Myoxini, whilst it is absent in all Insectivora except Macroscelis, Rhynchocyon, and the Tupaiidae. Both orders alike, it is true, have a discoid deciduate placenta, but in the Rodents the omphalo-mesenteric vessels take a large share in the nutrition of the foetus up to the end of pregnancy, the umbilical vesicle lines the whole of the chorion which is not occupied by the disk of the placenta proper; and this disk is never attached except to the mesometrial border of the uterine cornua; whilst in the Insectivora as in the Chiroptera the umbilical vesicle is attached, with the exception of Sorex, only over a limited circular area of the chorion opposite that occupied by the true placenta, the functional importance of its vessels is less, and the site of the allantoid placenta may be on any part of the uterine walls.