typical of the problems of jurisprudence. What significance is there in the statement (whether true or false) that dominium needs something more for its transfer than mere agreement, to a student who knows nothing of the relation between the Contract of Sale and the Conveyance in English law, or of that between Emptio and Traditio?

In the absence of conclusive à priori reasoning in favour of the study of jurisprudence apart from practical law, we are driven back upon experience, as to which there is not much recorded evidence. But the number of those who hold that jurisprudence is the proper introduction to the study of law is far larger than is that of those who, holding this view, base it upon experience or give any other reason for their belief. There is indeed one weighty champion, Professor Clark. But to his argument in the LAW QUARTERLY REVIEW (i. 204) it may be replied, that, while it is quite possible for the student to learn something about 'what subdivisions have been found useful,' it is very unlikely, unless he be of unusually good intelligence, that he will learn much as to the reply to the question 'upon what actual differences they depend.' If he do, it will be because he has gathered from the scattered references to actual systems more than the average student possibly could. There may indeed be a certain advantage in providing a student with a few elementary distinctions of which he does not understand the rationale, to serve as pegs on which to hang facts in a directed order, but these are no more than he will get in any set of lectures on Roman Law, and no more than he ought to get in any scientific text-book on any branch of any system of law. And by learning them in this way, he has a chance of ascertaining their rationale. There does not exist, and probably there never will exist, a text-book of jurisprudence with such a restricted aim as this. It certainly was not this aim which produced Professor Clark's searching criticism of Austin. This reference to a conceivable narrow scope for treatises on jurisprudence is not however altogether superfluous. For writers on jurisprudence are rather under suspicion of justifying to themselves the writing of books with the wider scope by recognising the fact that a book with the narrower scope might be written.

There remains another question. What success is likely to attend our search for these numerous common principles? The fact that most writers on jurisprudence are more concerned to show the differences than to point out the identities discoverable on comparison of systems is significant, and still more significant is the guarded way in which they speak of these identities. Professor Clark says, 'It is generally and not unreasonably assumed,

that there are certain principles,' &c., 'common to all systems of law' (L. Q. R. i. 203). Professor Holland omits all reference to this generality in his final definition (Jurisprudence, p. 12). Sir W. Markby is content to speak of those principles of law which are generally deemed universal.' One is surely entitled, without denying the existence of some universal principles, to regard with suspicion a science the exponents of which base it on an assumption of which they seem so much afraid. Austin is more decided. He divides these universal principles into those which are 'necessary,' and those which are not necessary.' It is not easy to gather the sense in which this word is used. Austin cites. several necessary' distinctions, with a clear intimation that there are many more, and yet scarcely one of those cited seems 'necessary,' unless this word be a synonym for 'obviously convenient.' It would be good to know, for instance, what claim the distinction between crime and civil injury has to this exalted rank. It may be answered that the aims of the two sets of rules are different

punishment is not redress. But it may be still asked whether a distinction is necessary' because it is real, whether jurisprudence is concerned with the aim of the legislator except in so far as the aim produces an effect upon legal rule, and whether the distinction between punishment and redress squares with this division of wrongs except upon a definition of punishment made in view of the division. Austin might have based his view on another ground, namely, the distinction between the rights of a subject and the 'quasi-rights' of the Sovereign. It is hard to accept Austin's proof of the proposition that State rights are not legal. There are, he says, essential to a legal right three persons: the Sovereign conferring the right, the person on whom it is conferred, and the person on whom the corresponding duty is imposed. In the case of state rights there are but two. In meeting the objection that there are all these three parts-though two of them are played by the same person -Austin seems to lose sight of his own analysis. In terms his argument is no more than a mere assertion. A man cannot, he says, confer rights on himself. To say he can is to confuse right and might. Here is, surely, that confusion between different senses of the word right, which he so carefully avoids elsewhere. A man's own dictum cannot make that act of his right which before was wrong. But Austin has shown clearly that the gist of a legal right is, that the person invested with it can have legal sanctions applied. These the Sovereign can himself apply, and the right so created differs from a mere power and will to do an evil, in that the evil is conditioned on a certain act or forbearance. What more can be said for any legal right? The Sovereign has not enlarged or contracted

his power; he has simply declared his intention as to the mode of exercise of a part of it, and has thus converted so much of it into an actual legal right. Austin's proof is the assumption that the three personae must reside in distinct individuals, an ordinary accompaniment of the differentia being in fact substituted for the differentia itself. Professor Holland rejects this view of Austin's on the ground that in current formulae the Sovereign is described as having rights (Jurisprudence, p. 110). But this fact, while it shows that Austin's distinction is disregarded in English practice, is in itself immaterial to the argument, since Austin might reply that the looseness of legal terminology does not alter the fact, if it be a fact, that the power vested in a Sovereign and called a right is not of the same nature as a legal right vested in the subject.

Similar objections may be taken to others of Austin's necessary notions. Similarity of terminology or even of rule does not imply identity of principle. If we limit ourselves to those cases where principle and rule based on it are the same in all systems, or even in the English and Roman systems, we shall find ourselves left with a very small residuum. Does the Roman division of actions into those in rem and in personam justify us in assuming that they would have distinguished between jura in rem and jura in personam on the lines we lay down? Would they have accepted Austin's distinction between servitus and dominium? Is the Roman conception Dominium really equivalent to ownership? The Roman distinctions, though similar to ours in general result, are based on different considerations. The real identities are yet to seek. Their persistence in text-books may be due to a conviction on the part of text-writers that the truths of jurisprudence appear more beautiful when shrouded in a little decent mystery.

Another conception which appears to be regarded as fundamental in jurisprudence is the notion thing.' In relation to this the same generalisation has led to much confusion. We are commonly taught that the Romans oscillated in their conception of a res between the physical thing and the rights over it. It seems better to break up the whole conception into three parts (as is done by Austin, p. 371, and by Professor Holland in an earlier edition of his work): the thing itself, the advantages which can arise out of it (acts and forbearances), and the right to those advantages, and to suggest that the very obvious confusion between the physical thing and the rights over it may be due to the existence of a still deeper confusion between the thing itself and the sum of the advantages it can offer. That the treatment of a right as a res is at bottom a confusion of distinct conceptions is a fact brought into strong relief by the inclusion of dominium among res corporales. It is noticeable

that the distinction between the res and the right to it is often brought out clearly, if implicitly, as in one of those pithy maxims in the last title of the Digest, Minus est actionem habere quam rem (D. 50. 17. 204), where actio of course means right of action.

6

The conception 'thing' appears in the analysis of the elements of a right into three-the persons concerned, the physical thing to which it relates, and the acts or forbearances which it contemplates. Austin conveniently marks off the last two as the subject and the object of the right. Professor Holland uses the word object where Austin uses subject, defining it as that over which the right is exercised' (Jurisprudence, p. 76). To this there is little objection. But he goes on to state the Roman view of a right as being a res, and to state it as a truth of general jurisprudence, taking his illustrations from different systems (p. 83). If rights are things, and things are definable as the objects of a right, then it follows that rights can be objects of a right, i. e. 'whatever is treated by the law as the object over which one person exercises a right, and with reference to which another person lies under a duty' (p. 83). It seems to follow indeed that every right which is a thing must be the object of a right. Professor Holland illustrates these definitions by ordinary physical things and by such res incorporales as a patent, an easement, hereditas, and he describes these as rights which for purposes of a transfer or otherwise are occasionally treated as if they were physical things' (p. 84). This remark follows from the attempt to make the Roman view of a right as a res square with the above. definition of a thing. It is perhaps unprofitable to enquire whether a right must be conceived of as a thing to make it transferable, though it seems more reasonable to say that rights, and rights only, are transferable in law- a view not inconsistent with the attitude adopted by both the English and Roman systems towards the transfer of a right without delivery of a res. Professor Holland (p. 87) speaks definitely of these res incorporales as cases where a right is the object of a right, adding, as another possible case, obligations. Whether such cases may arise or not, it is not now necessary to determine. But these at any rate are not instances. In the case of patents the 'object' as defined by the author is the inventor's new idea or, it may be, nothing (for there may well be rights that have no 'object' in this sense); in the case of easements it is the land; in the case of hereditas it may be a host of things. The patent right is no more the object of the right than ownership is owned. With as much reason a creditor's right may be defined as the right to exercise a right to recover a right, and out of any single right we can construct an unending series. Sir W. Markby states very shortly a view identical with that of Professor Holland. Is there

not here a real confusion, and is it not (though in part warranted by colloquial usage) the result of an attempt to incorporate into a system aiming at universality incompatible rules drawn from various systems?

Difficulties of this sort are likely to occur in any treatment of the subject, which at once admits and ignores the fact that law is based on conventions. This inconsistency results from a search for that unity which is described as lately discovered, but which seems to have been invented, and that not recently. It is true, of course, that the arrangement adopted by writers on jurisprudence is not alleged to be universal, except in the sense that it is universally applicable. It is the principles, notions, and distinctions,' the universality of which makes the general application of this arrangement possible. The fact that the arrangement adopted by one writer differs from that adopted by another is therefore immaterial. The fact that the principles of different systems while often similar are rarely identical, is no check to those who seek general truths, since they hold that a similarity of institutions proves the existence of a common principle which explains both forms. This is, at least, what Austin seems to mean when he speaks of certain distinctions as 'more or less nearly conceived in different systems' (p. 1108). This common element is the fact that human needs have been similar in different ages and among different peoples, a fact which is in itself no part of jurisprudence, though it bears upon that historical study which makes a scientific jurisprudence possible.

If we find ourselves unable to accept these very large assumptions which underlie the current philosophy of law, a question which naturally suggests itself is this. May we not find in comparative or historical jurisprudence (for to deserve either name it must be both) the philosophy of which we are in search? The answer seems to be that though this study may provide us with a philosophy of law, this philosophy will be in no way akin to Austin's 'General Jurisprudence.' It is impossible to dispute the enormous service rendered by historical study to the science of law. But the laws established by such writers as Sir H. Maine are laws of development. They show us how from rudimentary institutions similar in nature the most diverse rules have been developed in different systems. They point out connexions between matters apparently quite unrelated. The analogous biological studies provide scientists with systems of arrangement as well as laws of development, and we may fairly hope that comparative jurisprudence will do something in the same kind for lawyers. But its work is only just beginning, and no writer on jurisprudence can as yet base even his arrangement on the results obtained from