aggregate of rights pertaining to a person who has once acquired it and has not lost it. This confusion of possession as an actual physical relation and possession as an aggregate of rights constituted by a certain event and continuing until another event, runs right through the subject. A very good illustration may be found in the controversy as to possessio plurium in solidum.

We learn from Paulus (41. 2. 3. 5.) that Sabinus was of opinion that in the case of a tenancy at will both the precario dans and the precario accipiens had possession, and that Trebatius thought that two people could possess the same thing provided the possession of the one were justa and of the other injusta. Labeo and Paulus reject these views, the line of argument taken by the latter being as follows; non magis enim eadem possessio apud duos esse potest, quam ut tu stare videaris in eo loco, in quo ego sto, vel in quo ego sedeo, tu sedere videaris. This view appears to be approved by the compilers, but nevertheless a fragment of Pomponius is given in the Digest, which certainly is in agreement with the opinion expressed by Sabinus. Speaking of a precarium of a slave he says (43. 26. 15. 4.), placet autem penes utrumque esse eum hominem, qui precario datus esset, penes eum qui rogasset, quia possideat corpore, penes dominum, quia non discesserit animo possessione. And, again, the opinion of Trebatius given above is confirmed by Ulpian (43. 17. 3 pr.), Si duo possideant in solidum, videamus, quid sit dicendum.. ego possideo ex justa causa, tu vi aut clam: si a me possides, superior sum interdicto, si vero non a me, neuter nostrum vincetur: nam et tu possides et ego. There are great difficulties about this passage, but one thing is evident (unless we accept Savigny's idea that the final clause of this passage was intended as a reductio ad absurdum), namely, that Ulpian saw nothing impossible in possessio duorum in solidum.

This difference of opinion between the two contemporaries Ulpian and Paulus on such a point, shows us most plainly that they were arguing from two quite different conceptions of possession. Paulus (41. 2. 3. 5) treats possession as a mere fact, which is as completely independent of the influence of the judge or legislator as are the rules of arithmetic or the state of the weather. He says two people can no more possess the same thing than they can stand on the same spot. Ulpian, on the other hand, is thinking of possession as an aggregate of rights or legal consequences, which is in reality the creation of the legislator and which can be attributed to more persons than one in different degrees. The question with him is not-Is possession by two persons of the same thing physically possible? but-Can the jus possidendi be attributed to two people? or in other wordsIs the jus possidendi, after the model of the jus dominii, an exclusive

right, or, like the jus pignoris, a non-exclusive one? If the latter, then stronger and weaker rights of possession may be given to two persons over the same thing.

Sir F. Pollock, speaking of the same question with reference to the Common Law, says 'It must be admitted that the language of our authorities is anything but clear and uniform, and sometimes a bailor and bailee are spoken of as both having possession. In such passages the word is used in a double sense 1 This is just what happened in Roman law, but unfortunately by far the most important place in the Title de acquirenda vel amittenda possessione is given to Paulus, who showed more strongly than any jurist the inclination to treat possession as a mere fact, and who appears quite unconscious of the confusion wrought by the double sense of the word. Papinian seems less satisfied with the condition of the doctrine, but he, like Paulus, fails to distinguish between the two meanings of possession. Though Ulpian came nearer to the solution of the problem than either Paulus or Papinian he cannot be said to have solved it, and as the doctrine was left in this unsatisfactory condition by the last of the great jurists, nothing in the way of improvement could be expected from the imperial legislators, nor from Justinian's compilers.

Let us pass on now to another reproach which Mr. Justice Holmes has brought against the Roman law, namely its refusal of possessory remedies to bailees in general. It is in connection with this subject that I wish to speak of Ihering's latest contribution to legal science 2. The importance of this work reaches far beyond that of the question of possession, for it is, as its second title indicates, an attack on the method of German jurists from Savigny downwards. It is even more than this, for Ihering does not shrink from the attempt to trace the pernicious tendency, as he considers it, up to its source in the writings of the Roman jurists, of whom Paulus is shown to be the chief offender. He chose Possession as the ground on which the seed of error has brought forth its most luxuriant crops, and his work might more correctly be entitled 'The true and the false method in Jurisprudence, as illustrated by the doctrine of Possession.'

The manner in which this question of the denial of possessory remedies to bailees and tenants has been approached by different jurists, is in itself a study in juristic methods. Bruns and the other upholders of the Will theory explain the exclusion of bailees

1 Possession in the Common Law. Pollock & Wright, p. 21.

"Der Besitzwille. Zugleich eine Kritik der herrschenden juristischen Methode. Jena, 1889.

from possession by assuming a deficiency in their will: they accept the rules of the Roman law on this point as 'formulated reason,' and set to work to find a philosophical explanation for them. No one can be a 'true' possessor who recognises the ownership of another; when therefore the modern German law, like every other rational system, gives possessory remedies to the tenant and the bailee, they see in this a sacrifice of 'principle' to convenience. To which Mr. Holmes remarks in words which might be Ihering's, I cannot see what is left of a principle which avows itself inconsistent with convenience and the actual course of legislation. The first call of a theory of law is that it should fit the facts 1.'

If the Roman law of possession had come down to us in as fragmentary a condition as the XII Tables, and the German jurists were therefore free to indulge their taste for reconstruction, they might have given us a logical doctrine, though it would to a certainty have hopelessly broken down in practice. But there are those awkward cases of the pledge-creditor, the emphyteuta, the sequester and the precario tenens (to say nothing of more doubtful ones) to be explained, and the attempt to make them agree with preconceived theories has caused much difficulty. The Romans, with an indifference to the nature of true' possession, which is excusable as the Germans had not yet discovered it, gave them possessory remedies, although as they recognised the dominium of another, they certainly had not the animus domini. Savigny, as every one knows, set up for them the category of derivative possession,' which is nothing more than a collective term for the cases which his theory failed to explain2. He admits in short that they are anomalous and suggests the explanation of historical reasons.' But, as both Bekker and Ihering point out, there is no ground for assuming that the Romans regarded these cases as anomalous. The extension of an existing remedy on the ground of analogy was of very frequent occurrence, and the Roman jurists well understood the nature of these extensions and marked the remedies in question with appropriate names. Now in the cases of so-called derivative possession' there is no trace of any such distinction, and the only ground therefore for treating them as anomalous is their failure to fit into Savigny's theory. If the animus domini were presented as merely a rule admitting exceptions, it might still survive, though with a diminished lustre, but it is asserted to be essential to the very nature of possession3, and we are

1 Holmes, Common Law, p. 211.

2 Savigny is generally regarded as the originator of the theory of animus domini, and of derivative possession'; but Dr. Goldschmidt has pointed out that the former theory, in the identical form in which it appears in Savigny's work, is as old as Donellus, while 'derivative possession' may be traced to Lauterbach, a German jurist of the last century. Ihering, Besitzwille, pp. 246–252.

3

Savigny, Besitz, 109, 110.

therefore only left the choice between sacrificing it and throwing over the Roman law. And from this latter course some jurists have not shrunk. Puchta, a Roman lawyer himself be it noted, and not a professor of Naturrecht,' apparently considered possession and detention as conceptions existing from all time distinct from one another. The cases of derivative possession ought to have been detention, but an arbitrary jurisprudence decided otherwise. Puchta's objection to this classification is not, it need hardly be said, based on practical considerations: these are beneath the notice of the ideal jurist. According to the doctrine of possession written in the stars, the animus domini is essential, it is the soul which inhabits the corpus possessionis; if the Roman law gave possession in cases where this animus was wanting, so much the worse for the Roman law. Let the pledge-creditor, the emphyteuta and the rest, enjoy the practical advantages and even bear the name of possessors, they are not possessors in the true sense of the word for all that.

Bruns also is convinced that it belongs to the innere Nothwendigkeit of possession, to the eternal fitness of things in fact, that no one can really have possession who has not the animus domini1.

We seem to be listening to an echo of the Naturrecht of the cighteenth century, for this is a possession which never was on land or sea its dwelling-place is that abstract region 'wo die reinen Formen wohnen'; its creator is one who draws his legal principles not from actual life, but by logical deduction from premises of his own; the source of law in his sense is not the market-place and the law-court, but the study of the Professor. If we may judge a tree by its fruit, there must be a defect somewhere in the methods of the German historical school, for Puchta and Bruns are justly classed among its ablest representatives.

Savigny himself seldom forgot, as many of his disciples do, that his purpose was to expound the Roman law, and not to lay down. principles with which that law must conform at peril of professorial censure; but his attempt to save the animus domini by means of derivative possession' is like the act of a man who having sent out an unseaworthy ship of his own making, sends out with it a leaky boat in case of accidents.

But though Savigny to the last clung to the animus domini, the impossibility of making it square with the facts of Roman law is more and more generally recognised, and the animus possidendi, defined as the intent to exercise exclusive control over the thing for oneself, substituted for it. This is an improvement on Savigny's view, for it takes away the anomalous character of the instances of so-called derivative possession, but what in Ihering's Bruns, Recht des Besitzes, u. s. w., 463-471.

opinion is the fundamental mistake of assuming any specific intent for possession distinct from that required for detention still remains.

Why had the emphyteuta and the pledge-creditor the possessory interdicts, whereas the tenant for years and the hirer had them not? The usual answer given is, that the former had the animus domini or animus possidendi, while the latter had not the intent to possess for themselves, but for another. Ihering, on the other hand, declares that the intent of the parties has nothing whatever to do with the distinction; that the intent which is a necessary constituent of every legal act, is the same in the case of both possessor and detentor. Wherever the fact of possession exists there is legal possession, unless a rule of law exceptionally enacts that there shall be only detention. The 'Besitzwille' or animus possidendi never comes into question in the decision of an actual case, and a judex might therefore apply the Roman law of possession correctly without knowing anything about the theories of the will. On Ihering's theory the task of the judex is simple: his instructions are-assume legal possession wherever the fact of possession exists, unless one of the causae detentionis, which you will find enumerated in any treatise on the Pandects, is proved. The causae detentionis are positive rules of law, based in all cases on practical utility. They are objective rules, that is to say, they rest on considerations outside the person of the detentor, and not, as the prevalent doctrine teaches, on some deficiency in his will.

The impossibility of distinguishing between possession and detention, if the animus be taken as the test, may be shown by an example. Let us take the agent and tutor, and, on the other hand, the tenant. The two former, according to the Will theory, should have merely detentio, as they intend to possess for another; the latter should have possession, as he intends to possess for himself: the jurist of the Augustan age would, however, have given possession to the two former and detention to the latter. The conclusion Thering draws is that, it being impossible to deduce logically the distinction between them, we must seek the explanation from history.

He finds the origin of detentio or possessio naturalis in the relation of those under power to the paterfamilias. A filiusfamilias in early Roman law could not be an owner, nor could he bring actions: he could not, therefore, be legally a possessor. But the institution of peculium was a very general one in the case of sons of full age. In rich families the sons had frequently very large independent establishments of their own, with country estates and numerous slaves, while among the trading classes they often had a separate