supplies a very striking counterpart to this, and justifies my reference to it, even if it should turn out to have no relation to the amalaka of the sikhara.1

Here is what Col. Marshall discovered regarding this stone, he first learned that there were relics in the boath. He and his friend tried to get in to see them, but the Toda in authority would on no account allow this; the place was far too sacred. So they determined on a midnight expedition to the temple, with the intention of getting the desired knowledge in a burglarious manner. A very humorous account of the adventure is given. The two gentlemen got in all right, but, to their sad disappointment, found nothing that they expected; only some ordinary articles were in the place,-"no bell, no hatchet, neither ring nor relic of any kind, no niche for lights, no altar, no stone, no phallus or lingam. No snakes! Every one has been telling us lies, and the world is full of sawdust."2 As the old Toda on whose information they depended had always given correct information before, he was cross-questioned a day or two later. After some preliminary inquiries, he was asked where the relics were placed, and, "with his hand to the side of his mouth, he said in a low voice, 'Under the stone on the top of the roof." "3 In the case of the Todas, these, whatever they were, are not necessarily human relics. The point is that they were relics—something sacred to be preserved— and that is motive enough when we are seeking for origins in architecture. Now it must be evident, if this, which we may easily believe is a very primitive sort of temple, had a relic on the summit of its spire, and the dagoba had a relicholder on its top, it adds considerably to the probability that the amalaka is a form derived from something of the same kind. Here, again, we have no direct proof: it may have been so, or it may not. The suggestion may be useful as a hint to others, but it must remain a suggestion only, till further knowledge has been obtained.

The Todas cremate their dead, and they have two burn

1 Note also the upper half of roof of Bengal temple. Here this particular feature is very strongly marked.

2

p. 166.

3

p. 167.

ings. Colonel Marshall and Mr. Breeks seem to me to differ slightly in their accounts, but I shall follow the latter. After death the body is burnt, but the skull is preserved; also a portion of a finger-nail, cut off, I suppose, before the burning. These are kept for about a twelvemonth, and then they are burned with a number of articles. The burning is done at a stone circle,1 and at the entrance a hole is made in the earth, into which the ashes are placed; a stone is laid over them, and a man breaks a chatty over the stone. This part of breaking the chatty is a custom followed more or less by all the primitive tribes of the Nilgiris. I give this account because if this stone with the ashes under it has any connection with the origin of the stone, similarly with relics under it, on the summit of the boath, we have here what might be the explanation of the Kalasa, or vase, which surmounts the amalaka on the Hindu sikhara. This is, of course, assuming the suggestion given above regarding the amalaka is correct.2

I cannot help suspecting that the Toda customs represent at the present day a very primitive condition of the Hindu rites, or perhaps I ought rather to repeat Mr. Fergusson's expression of "Dasya rites." I am not sure whether the bell figures in the old Vedic ceremonies, but we know that it does so very largely in the worship of Siva now. All his temples have a bell, which is sounded by the worshippers, and Nandi has always one hanging from his neck. With the Todas a bell is the most sacred relic in the temple. It is supposed to be old, and has no tongue; a bell is always placed round the necks of the buffalos sacrificed at the cremation; the relics which are preserved from the first burning

1 Mr. Breeks states that the Toda burning-place is called "Methgudi, lit. Marriage Temple," p. 20. This suggests an explanation of the Asura festal rites in relation to the dead.

2 In many Himalayan sikharas, instead of the amalaka there is a small roof formed of wood; it is square, and a pyramid in shape, standing on four small wooden posts. This very marked variation is, I think, a point in favour of the theory that the amalaka is derived from an umbrella, which would be like the wooden structure and canopy. I believe some of the Himalayan temples have more than one of these roofs, one above the other, in this again still more suggesting the umbrellas of the Buddhist dagoba, which Mr. Fergusson believes to be the source of the sikhara.

are placed in a hut, and a bell is hung over them, which the relatives ring night and morning, generally for nearly a year, when the second cremation takes place. When the votary of Siva at the present day rings a bell at a shrine, which he supposes is to waken or to call the attention of the god, he may be only repeating part of an old rite connected with the dead, of which we have a marked example in the "dead bell" of the Roman Catholic Church. The Hindu of our own time will not kill a cow, in later times he has adopted a more humane ritual; but his Nandi may yet represent the old funereal sacrifice which accompanied the spirit of his proprietor, and was thus a sort of Vahan, to the regions of Yama.?

I can refer to a noted bell of this kind which existed in Glasgow, and was said to have belonged to St. Mungo, the patron saint of the town; it was known as the "Deid Bell," and was used at funerals; it "was also rung through the streets for the repose of the souls of the departed." This bell even survived the destruction of many things at the Reformation, as the following record of a Presbytery meeting in 1594 will show: "The Presbyterie declairis the office of the ringing of the bell to the buriall of the deid to be ecclesiasticall, and that the electioun of the persone to the ringing of the said bell belongis to the ancient canonis and discipline of the reformit kirk." This bell still survives, but only in the armorial bearings of the city.

The Vahan of Yama is curiously enough a buffalo, the animal sacrificed at the Toda cremations.

72

ART. III.-The Chaghatai Mughals. By E. E. OLIVER, M.I.C.E., M.R.A.S.

WITHOUT attempting to go back to the obscure traditions concerning the great nomad confederacy or confederacies that ranged the country north of the desert of Gobi, or to the genealogies of the tribes of Turks, Tārtārs, and Mughals, descendants of Yafiṣ (Japhat) son of Nūḥ, who, after coming out of the Ark with his father, is said to have fixed his yūrat or encampment in the Farther East, and who have furnished subjects for the most copious traditions for native chroniclers, and materials for the most intricate controversies ever since; it may perhaps safely be assumed that Mughal was probably in the first instance the name of one tribe among many, a clan among clans, and extended to the whole as its chief acquired an ascendency over the rest. The name is most likely locally much older than the time of Chengiz, but it was hardly known to more distant nations before the tenth century, and became only widely famous in connection with him.

It is also perhaps unnecessary to enter upon the vexed question as to how the name is to be most properly spelt. Writers who have drawn considerably from Chinese sources, and most of the standard authors, like d'Ohsson, Yule, Howorth, and others, have adopted and familiarized us with "Mongols." On the other hand, to the Persian writers who have much to tell concerning them, and in so far as they are associated with India and the countries adjoining, they are Mughals or Mughūls. To Timur, Báber, and Akbar, their ancestors were Mughals, and the first "Irruptions of