distinguished offspring, and he will not die childless.' But if he performs it on the first day of the halfmonth, the caution is given that the issue of the sacrificer will consist chiefly of daughters. When, however, we come to writers of a much later era, like Vishnu, we find a distribution of the sacrifices which is very significant. Vishnu gives us a summary of the whole ritual of ancestor-worship as practised at the date of the treatise called by this name (Vishnu, chap. lxxiii.) First of all the sacrificer is to worship the (greater) Gods. Then on particular days-the ninth days of the dark halves of certain months-he is to consecrate an offering with proper hymns and scriptural texts and present it to three Brahmans present, who represent his father, grandfather, and great-grandfather. liturgy and ritual which he is to follow are indicated, head by head, and it is essential for the virtue of the sacrifice that a company of Brahmans should have been invited. On certain other sacred days, the Anvashtakas, he is to sacrifice to his mother, his paternal grandmother, and his paternal great-grandmother; and lastly, says the writer, an intelligent man'-an expression which, as it appears to me, is always used of a doubtful point-' must offer shraddas to his maternal grandfather, and to the father and grandfather of him in the same way.' The order of celebration seems to me to follow the historical order,

The

and to show that the ancestors first worshipped by the Hindus were the father, grandfather, and greatgrandfather.

It is clear then, I think, that wherever ancestorworship arose, Paternity was fully recognised; and as the texts relating to this worship are as old as any others in the sacerdotal law-books, and indeed are probably the oldest, I attach small importance to casual expressions found here and there in these treatises which have been thought to show that their writers preserved traditions of the savage custom of tracing descent through females only. Still, as we cannot doubt the existence and prevalence among some part of mankind of this savage usage, sometimes called 'Mother-law,' it is impossible not to ask oneself the question, Did the worship of the dead bring about the recognition of paternity, or is ancestor-worship a religious interpretation of, or a religious system founded upon, an already existing institution? M. Fustel de Coulanges, without referring to the custom of Motherlaw,' certainly seems to me to express himself occasionally as if he thought that all the characteristics of the so-called Patriarchal Family were created by the worship of ancestors which was ever celebrated in the recesses of the household; and that from this worship sprang the Father's Power as its high-priest, and also the denial of kinship to persons no longer

able to participate in it, as the married daughter and the emancipated son. It may well be believed that ancestor-worship, by consecrating, strengthened all family relations, but in the present state of these inquiries the evidence certainly seems to be in favour of the view that the Father's Power is older than the practice of worshipping him. Why should the dead Father be worshipped more than any other member of the household unless he was the most prominent—it may be said, the most awful-figure in it during his life? It was he, according to the theory which I have described, who would most frequently show himself, affectionate or menacing, to his sleeping children. This opinion is fortified by the recent investigations into the customary law of the Punjab, the earliest Indian home, I must repeat, of the Aryan Hindus after their descent from the mountain-land of their origin. Ancestor-worship does exist among the Hindus of the Punjab. But it is a comparatively obscure superstition. It has not received anything like the elaboration given to it by the priesthood in the provinces to the south-east, many of whose fundamental doctrines are unknown to the Punjabee communities of Hindus. Nevertheless, the constitution of the Family is entirely, to use the Roman phrase, 'agnatic;' kinship is counted through male descents only. There is a very strong resemblance between these usages and the most ancient Roman law, and their differences, where they differ, throw very

valuable light on the more famous of the two systems.

The truth seems to be that, although Ancestorworship had at first a tendency to consolidate the ancient constitution of the Family, its later tendency was to dissolve it. Looking at the Hindu system as a whole, we can see that, as its historical growth proceeded, the sacerdotal lawyers fell under a strong temptation to multiply the persons who were privileged to offer the sacrifices, partly in the interest of the dead ancestor, chiefly in the interest of the living Brahman. In this way, persons excluded from the ancient family circle, such as the descendants of female kinsmen, were gradually admitted to participate in the oblations and share in the inheritance. Some traces of a movement in this direction are to be found throughout the law-books; and a very learned Indian lawyer (Mr. J. D. Mayne, 'Hindu Law and Usage,' chap. xvi.) has shown that, wherever in modern India the doctrine of Spiritual Benefit-that is, of an intimate connection between the religious blessing and the civil right of succession-is most strongly held, women and the descendants of women are oftenest permitted to inherit. It is remarkable that the Equity of the Roman Prætor, which was probably a religious before it was a philosophical system, had precisely the same effect in breaking up the structure of the ancient Roman family, governed by the Father as its chief.

CHAPTER IV.

ANCESTOR-WORSHIP AND INHERITANCE.

THE close connection between succession to property after death and the performance of some sort of sacrificial rites in honour of the deceased has long been known to students of classical antiquity. A considerable proportion of the not very plentiful remains of Greek legal argument to be found in the Athenian Orators is occupied with questions of inheritance, and the advocate or litigant frequently speaks of the sacrifices and the succession as inseparable. Decide between us,' he says, 'which of us should have the succession and make the sacrifices at the tomb' (Isæus, 'In the goods of Philoctemon,' Or. vi.) 'I beseech you by the gods and immortal spirits not to allow the dead to be outraged by these men; do not suffer his worst enemies to sacrifice at his grave' (Or. ii.). In a former work I pointed out the number, costliness, and importance of these ceremonies and oblations among the Romans, and I insisted on their probable significance as the source