needs to be explained. But for archaic society the collective body and not the individual is the natural person.

We find the same conditions existing in full force among the German tribes in a much later period of time than that which Maine is directly considering in this chapter. A recent learned writer in France, dealing with precisely the same subject as it occurs in the mediaeval history of French law, has forcibly contrasted the Roman conception, as it was established in the classical law of the empire, with the German.

'Le droit romain consacre le triomphe de l'individualisme; la volonté personnelle du chef de famille, voilà le facteur juridique essentiel, l'agent de toutes les transactions, la force créatrice de tous les droits. Cette volonté est si respectée et si puissante, qu'elle continue d'agir après la disparition de celui qui l'a exprimée. Le père règle le sort de sa fortune et de sa famille pour le temps où il ne sera plus, et cela par un acte souverainement libre, qu'il est toujours à même de modifier. . . . L'individu sui iuris est, dans le monde romain, l'unité juridique et sociale.

Chez les Germains, c'est bien plutôt la famille. Il serait sans doute excessif, surtout pour le temps des Leges [the custumals collectively known as "Leges Barbarorum"], de déclarer en termes absolus que la famille est tout et que l'individu n'est rien; la vérité sous cette forme serait exagérée et dénaturée. Mais il est certain cependant que l'exaltation de l'individu est beaucoup moins complète qu'à Rome, et que d'autre part la famille forme une association, une sorte d'être collectif armé de droits inconnus des jurisconsultes de l'Empire. L'énergie individuelle est limitée dans le temps, et les Germains ne peuvent pas concevoir qu'elle s'exerce au delà de la tombe; sitôt l'homme mort, toutes ses volontés s'évanouissent. Au même moment ses prérogatives juridiques sont recouvertes et absorbées par celles de ses parents, car de son vivant même sa famille jouissait de droits autonomes qu'il ne dépendait pas de lui de supprimer: sa mort les développe, mais elle ne les crée pas.' (Auffroy, Evolution du testament en France, Paris, 1899, pp. 173-4. Cf. Brunner, Grundzüge der deutschen Rechtsgeschichte, 56: Das germanische Erbrecht war ein Familienrecht. For examples of analogous customs among various uncivilized tribes, see Lord Avebury, Origin of Civilisation, 6th ed. pp. 489-91.)

The suggestion in Maine's text of regarding the Roman ancestor in his representative character as a kind of corporation sole may be helpful to English students, but we can hardly trust it to throw light on the actual formation of Roman legal ideas. For our English category of corporations sole is not only, as Maine calls it, a fiction, but modern, anomalous, and of no practical use. When a parson or other solely corporate office-holder dies, there is no one to act for the corporation until a successor is appointed, and, when appointed, that successor can do nothing which he could not do

without being called a corporation sole. In the case of the parson even the continuity of the freehold is not saved, and it is said to be in abeyance in the interval. As for the king, or 'the Crown,' being a corporation sole, the language of our books appears to be nothing but a clumsy and, after all, ineffective device to avoid openly personifying the State. The problems of federal politics in Canada and Australia threaten to make the fiction complex. Is 'the Crown' a trustee for Dominion and Province, for Commonwealth and State, with possibly conflicting interests? or is there one indivisible Crown being or having several persons for different purposes? (F. W. Maitland, L. Q. R. xvi. 335, xvii. 131; W. Harrison Moore, L. Q. R. xx. 351; Markby, Elements of Law, § 145)1. The whole thing seems to have arisen from the technical difficulty of making grants to a parson and his successors after the practice of making them to God and the patron saint had been discontinued, as tending to bring the saints into the unseemly position of litigants before secular courts. All this, we may now think, makes for historical curiosity rather than philosophical edification.

But in any case the chief part of Maine's argument, his insistence on the theory of a man's posthumous existence in the person of his heir,' and the intimate connexion of that theory with the ancestor's representative character as head of the family, goes to the root of the matter. Mr. Justice Holmes, now of the Supreme Court of the United States, writing twenty years after Maine, summed this up with concise elegance (The Common Law, p. 343):—

'If the family was the owner of the property administered by a paterfamilias, its rights remained unaffected by the death of its temporary head. The family continued, although the head died. And when, probably by a gradual change, the paterfamilias came to be regarded as owner, instead of a simple manager of the family rights, the nature and continuity of those rights did not change with the title to them. The familia continued to the heirs as it was left by the ancestor. . . .

The aggregate of the ancestor's rights and duties, or, to use the technical phrase, the total persona sustained by him, was early separated from his natural personality. For this persona was but the aggregate of what had formerly been family rights and duties, and was originally sustained by any individual only as the family head. Hence it was said to be continued by the inheritance; and when the heir assumed it, he had his action in respect of injuries previously committed.'

Maine proceeds to trace the development of the Roman testament from a distribution of property, taking effect at once, made in

See too Williams v. Howarth [1905] A. C. 55', noted p. 6 above.

contemplation of impending death or great peril, and requiring, in its earliest form, something like legislative sanction (ep. Girard, Manuel, pp. 792-5), through the intermediate stage of a conveyance reserving a life interest, which may be seen in the provincial customs of the Roman Empire, and much later in mediaeval and even modern systems. Muirhead (Historical Introduction to the Private Law of Rome, pp. 66, 168) pointed out a remedy for the difficulty suggested by Maine, that a will by mancipation must have left the testator penniless. Usufruct might very well be reserved on a mancipation, Gai. ii. 33, 'and a reservation of a life interest in one's own familia would possibly be construed even more liberally than an ordinary usufruct.' Still, usufruct is not among the earliest institutions, and it would be rash to say that the difficulty may not have been real at one time. But men have been driven all over the world, by an imperfect state of property law or by special reasons for avoiding publicity, to put very large trust in the honour of chosen friends and assistants; and there is nothing about the Roman familiae emtor in the most archaic stage to surprise an English student who has made acquaintance with our mediaeval feoffee to uses. Indian practice will furnish a parallel in the benámí (literally, 'anonymous') conveyances to a nominal purchaser, to hold on a secret trust for the real one, which appear to have survived the original reasons for them. Sohm, however, holds (Institutes, § 99, p. 569, in Ledlie's translation, 2nd ed.) that the testament per aes et libram was coupled with a mandate to the familiae emtor, which was binding under the well-known provision of the Twelve Tables, 'uti lingua nuncupassit ita ius esto.' This would of course simplify the matter. The same learned author's suggestion that the institution of an heir was a modified form of adoption-that is, an adoption deferred to the testator's death-does not seem to be generally accepted (Girard, Manuel, p. 793).

What is said in this chapter about Hindu law would no doubt have been fuller if a convenient and trustworthy textbook like Mr. Mayne's had existed at the time when it was written. I am not aware, however, that any modification is needed except on one point, namely that the strict determination of the order of succession among an ancestor's next of kin according to the spiritual efficacy of their sacrifices is found only in the school of Bengal, and is thought to be a deliberate Brahmanical innovation (see, however, as to probable Buddhist elements Mr. Justice Sarada Charan Mitra's papers in the L.Q. R. for Oct. 1905, p. 380, and in the present number). As Maine himself said in 1883,' we now can discern something of the real relation which the sacerdotal Hindu law bears to the true

ancient law of the race' (Early Law and Custom, p. 194; see also the chapter on Ancestor-Worship and Inheritance). The general importance of keeping up the family ritual both in Hindu and in other archaic law remains undoubted. Some addition has to be made as regards the Hindu will. Quite unknown to early Hindu law, will-making came into use in modern times, though not in imitation of European practice according to the best authorities, and was not recognized in any of the Presidency Courts before 1832, when it was allowed in Bengal. When 'Ancient Law' was published the law was not yet quite settled in Madras and Bombay; but the courts of those Presidencies followed the same course within a few years. Apparently the first form of the Bengal will was a gift mortis causa to religious uses. The reader will perceive the resemblance to the development of the testament of chattels, under ecclesiastical influence, in mediaeval English law. The English history, however, is for the most part too complex and peculiar to throw much light on the normal type of evolution. As for the Anglo-Saxon will, even if it can be assimilated to modern wills, which is doubtful, it was a special and anomalous kind of document, and disappeared after the Norman Conquest. Probably language is still to be found in popular books asserting or implying that before the Conquest there was general freedom of alienation; but this is due to pure misunderstanding, the privileged class of transactions which are recorded in the AngloSaxon charters having been taken as typical and indigenous. Early English 'post obit gifts' (Pollock and Maitland, H. E. L. ii. 517 sqq., and see Note Q below) do present some analogy to the Roman will by mancipation; and this appears in a strengthened form in the conveyance to feoffees to uses to be declared by the feoffee's will which was common in the later Middle Ages. In the thirteenth century divers learned clerks made an ingenious and, it seems, almost a successful attempt to create posthumous disposing power by grants inter vivos, containing in what we now call the 'habendum' such words as 'cuicunque dare vel etiam legare voluerit.' A clause so framed is quite common in deeds of the third and even fourth quarters of that century, and inconsistent utterances in Bracton show that learned opinion fluctuated (186, 412b, pro, 49a, fuller and seemingly more deliberate, contra, cp. Pollock and Maitland, ii. 27). We may believe1 that for some time and to some extent the power such clauses purported to confer was exercised without objection. But this was a transitory

1 Extant wills of the period which purport to devise parcels of land (Madox, Form. Anglic. DCCLXVIII, DCCLXIX, DCCLXXI) are not conclusive as to the practice in the absence of a known previous grant with which they can be connected, as other explanations are possible.

experiment, and has nothing to do with any real testamentary distribution or succession. Local customs to devise land or, at any rate, purchased land existed, but their origin and early history are still obscure.

In Scotland we find the most remarkable illustration of the prae-testamentary stage, as we may call it, of property law. Properly there is no such term as Will in Scots law, and there was no true will of lands before 1868. 'Heritage could only be transmitted by a deed containing words of de praesenti disposition, and the use of the word "dispone" was essential' (Green's Encycl. of the Law of Scotland, s. v. Will). The accustomed form was (and apparently still is, notwithstanding that it is no longer necessary) a trust disposition and settlement,' a present conveyance reserving a life interest to the grantor. Scotland, in fact, is the last home of the old Germanic Vergabung von Todes wegen (Goffin, The Testamentary Executor, 1901, pp. 19, 99). It may survive many generations yet, for aught we know, as in the customs of Egypt and other parts of the Roman Empire essentially similar forms continued in use long after true wills had become familiar in the law of Rome. Original examples of the second century A.D. found at Naucratis might be seen in London some years ago. Notwithstanding the marks of Roman influence which the modern. English will bears, its practical scope and effect remain as different as possible from those of the Roman testament. As a rule the wills of Englishmen having any considerable property to dispose of aim not at investing any one person with the whole of the testator's control over his estate, subject to payment of debts and legacies, but rather at postponing absolute control and preserving the estate under the sanction of a trust which will not be finally determined while any child of the testator is a minor or his widow living. The capital is to be intact as long as possible, while the income is enjoyed or applied according to the testator's directions. If any one is at all like a Roman heir, it is the executor, who does not necessarily take any beneficial interest, and whose origin is quite different (Goffin, op. cit. p. 33; O. W. Holmes, L. Q. R. i, 165-6; Gierke, Grundzüge des deutschen Privatrechts, § 126, in Encykl. d. Rechtswiss. i. 555). The Roman horror of intestacy mentioned in the early part of the following chapter was equalled or surpassed among mediaeval Englishmen (Pollock and Maitland, ii. 356); but the reason was not one that would have occurred to any Roman from the time of Labeo to that of Justinian, being the danger to the intestate's soul if he died without having assigned a fitting part of his estate to pious uses (Du Cange, s.v. intestatio).