Regarding the shedding of these horns, it is supposed that the operation is connected with the sexual functions. It is a curious fact that castration has a powerful effect on this operation; if done early no horns appear; if later in life, the horns become persistent and are not shed.

Captain James Forsyth (in his 'Highlands of Central India'), was of opinion that the Sambar does not shed its horns annually, and states that this also is the opinion of native shikaris in Central India. This, however, requires further investigation. I certainly never heard of such a theory amongst them, nor noticed the departure from the normal state.

There have been several classifications of the Cervidæ, but I think the most complete and desirable one is that of Sir Victor Brooke (see 'P. Z. S.' 1878, p. 883), which I shall endeavour to give in a condensed form. Dr. Gray's classification was based on three forms of antlers and the shape of the tail. But Sir Victor Brooke's is founded on more reliable osteological details. As I before stated in my introductory remarks on the Ruminantia, the first and fourth digits, there being no thumb, are but rudimentary, the metacarpal bones being reduced to mere splints; the digital phalanges are always in the same place, and bear the little false hoofs, which are situated behind and a little above the large centre ones, but the metacarpal splint is not always in the same place; it may either be annexed to the phalanges, or widely separated from them and placed directly under the carpus. The position of these splints is an important factor in the classification of the Cervidæ into two divisions, distinguished by Sir Victor Brooke as the Plesiometacarpals, in which the splint is near the carpus, and the Telemetacarpals, in which the splint is far from the carpus, and articulated with the digital phalanges. All the known species of deer can be classified under these two heads; and it is a significant fact that this pedal division is borne out by certain cranial peculiarities discovered by Professor Garrod, and also, to a certain extent, by an arrangement of hair-tufts on the tarsus and metatarsus. In the Old World deer, which are with few exceptions Plesiometacarpi, those which have these tufts have them above the middle of the metatarsus, and those of the New World, which are, with one exception, Telemetacarpi, have them, when present, below the middle of the metatarsus.

There is also another character in addition to the cranial one before alluded to, which was also noticed by Professor Garrod. The first cranial peculiarity is that in Telemetacarpi, as a rule, the vertical plate developed from the lower surface of the vomer is prolonged sufficiently downwards and backwards to become anchylosed to the horizontal plate of the palatals, forming a septum completely dividing the nasal cavity into two chambers. In the Plesiometacarpi this vertical plate is not sufficiently developed to reach the horizontal plate of the palatals. The second cranial peculiarity is that in the Old World deer (Plesiometacarpi), the ascending rami of the premaxillæ articulate with the nasals with one or two exceptions, whereas in the New World deer (Telemetacarpi), with one or two exceptions, the rami of the premaxillæ do not reach the nasals. It will thus be seen that the osteological characters of the head and feet agree in a singularly fortunate manner, and, when taken in connection with the external signs afforded by the metatarsal tufts, prove conclusively the value of the system. In India we have to deal exclusively with the Plesiometacarpi, our nearest members of the other division being the Chinese water-deer (Hydropotes inermis), and probably Capreolus pygargus from Yarkand, the horns of a roebuck in velvet attached to a strip of skin having been brought down by the Mission to that country in 1873-74.

Now comes the more difficult task of subdividing these sections into genera—a subject which has taxed the powers of many naturalists, and which is still in a far from perfect state. To all proposed arrangements some exception can be taken, and the following system is not free from objection, but it is on the whole the most reliable; and this system is founded on the form of the antler, which runs from a single spike, as in the South American Coassus, to the many branches of the red deer (Cervus elaphas); and all the various changes on which we found genera are in successive stages produced in the red deer, which we may accept as the highest development; for instance, the stag in its first year develops but a single straight "beam" antler, when it is called a "brocket," and it is the same as the South American brocket (Coassus). On this being shed the next spring produces a small branch from the base of this beam, called the brow antler, which is identical almost with the single bifurcated horn of the Furcifer from Chili. The stag is then technically known as a "spayad." In the third year an extra front branch is formed, known as the tres-tine. The antler then resembles the rusine type, of which our sambar stag is an example. In the fourth year the top of the main beam throws out several small tines called "sur-royals," and the brow antler receives an addition higher up called the "bez-tine." The animal is then a "staggard." In the fifth year the "sur-royals" become more numerous, and the whole antler heavier in the "stag," whose next promotion is to that of "great hart" of ten or more points. The finest heads are found in the German forests. Sir Victor Brooke alludes to some in the hunting Schloss of Moritzburg of the 15th to 17th century, of enormous size, bearing from 25 to 50 points—50 inches round the outside curve, 10 inches in circumference round the smallest part of the beam, and of one of which the spread between the coronal tines is 74 inches. Professor Garrod mentions one as having sixty-six points, and states that Lord Powerscourt has in his possession a pair with forty-five tines. The deer with which we have to deal range from the elaphine, or red deer type, to the simple bifurcated antler of the muntjac, which consists of a beam and brow antler only. We then come to the rusine type of three points only—brow, tres, and royal tines, and of this number are also the spotted and hog deer of India, but the arrangement of the tines is different; and following the rusine type comes the rucervine, in which the tres and royal tines break out into points—the tres-tine usually bifurcate, and the royal with two, three or more points. The arrangements of the main limbs of the horns is strictly rusine—that is to say, the external and anterior tine is equal to or shorter than the royal tine, whereas it is the reverse in the axis (spotted deer), and therefore this genus should come between the two. Even in the sambar and axis there is a tendency to throw out abnormal tines. There are many examples in the Indian Museum, and I possess a magnificent head which bears a large abnormal tine on one horn, and a faint inclination in the corresponding spot on the other horn to do likewise. I have no doubt, had the animal lived another year, the second extra tine would have been developed. Professor Garrod has three phases of the rucervine type, which he calls the normal, the intermediate, and the extreme. The first has both branches of the beam, tres and royal of equal size (ex. Schomburgk's deer); the second has the tres-tine larger than the royal (ex. our swamp deer); and the extreme type is that in which the royal is represented merely by a snag, the whole horn being bent forward (ex. the Burmese Panolia Eldii). The true cervine type of horn I have already described in its progress from youth to age. The Kashmir and Sikim stags are the representatives of this form in India. In Japan there is an intermediate form in Cervus sika which has no bez-tine.

Stag with Horns matured.

Deer have large eye-pits, but no groin-pits; feet-pits in all four, or sometimes only in the hind feet. The female has four mammæ.

Stag with Horns in velvet.