nothing to be desired. Its excellent arrangement serves admirably to bring out the law in its historical development, and, notwithstanding its brevity, its fulness of detail, aided by an adequate tabular analysis, will render it invaluable for a hasty review.

TRADE UNION LAW AND CASES. By Herman Cohen and George Howell, F. S. S. London: Sweet and Maxwell, Limited. 1901. pp. xiii, 250. During the last thirty years the law of England as to trade unions and their members has been greatly changed by Acts of Parliament. Trade unions have become for many purposes legally recognized organizations with legal rights and subject to legal control, strikes are no longer illegal as "in restraint of trade," and the decision in Allen v. Flood has practically made the civil action for conspiracy of no effect as against trade unions. It is to enable the workingman readily to ascertain what the law on such matters now is that the present work has been written. The first and introductory chapter states briefly the history and effect of the seven more important Acts of Parliament as to trade unions. In the rest of the book is given the text of these acts, to which is appended copious annotation under the various sections. The work of the authors appears almost entirely in these annotations, which consist largely of quotations from the decisions of the courts. The collection of cases cited numbers over one hundred, and undertakes to be exhaustive. As the authors state, the book is not intended to be a legal treatise, but rather a working guide and manual for any one who has occasion to know and act on the present English law as to trade unions. For this purpose it would seem that the book must be of considerable value to the English trade unionist, but it is obvious that its field of practical utility must be confined to Great Britain.

We have also received:-
:-

THE LAW AND PROCEDURE OF UNITED STATES COURTS. By John W. Dwyer, LL. M. Ann Arbor: George Wahr. 1901. pp. xxi, 339. Review will follow.

GENERAL DIGEST, AMERICAN AND ENGLISH. Bi-monthly advance sheets, February, 1901. Rochester Lawyer's Coöperative Publishing Co. 1901. pp. 770.

HARVARD

LAW REVIEW.

VOL. XV.

JUNE, 1901.

No. 2

A

REQUIRED NUMBERS OF WITNESSES; BRIEF HISTORY OF THE NUMERICAL SYSTEM IN ENGLAND.

IT

T is well known that in the civil law of Continental Europe, the great rival of the English common law system, the process of proof rested fundamentally on a numerical system, according to which a single witness to a fact was in general not sufficient, specific numbers of witnesses were in certain cases required, and in some regions, and for some purposes, the weight to be given to each witness' testimony was measured and represented in numerical values, even by counting halves and quarters of a witness; and this system continued in force down to comparatively recent times. In the English common law system of jury trial, on the other hand, it was completely otherwise. At common law there was but a single instance, and that a borrowed one, of almost accidental and of entirely anomalous origin (the rule in perjury), in which a numerical rule existed; what little else there is to-day of that sort has come into our system either by express statutes (all but one dating since 1800), or by the filtration of civil law rules through the court of chancery, or by local judicial invention. The reason of this contrast, and of our successful resistance to the civil law rules, and the causes of our freedom from a principle of evidence now generally acknowledged to be unsound and deleterious, form a history worth examining.

I. The Numerical System in general.

(1) It has been doubted whether the Roman law in its prime (that is, before 300 A. D.) proceeded upon a numerical system in its treatment of witnesses. But it is clear that by the time of the Emperor Constantine, and also in the later codification of the Emperor Justinian, which served as a sufficient foundation for the Continental civil law, the Roman law had adopted the general rule that one witness alone was insufficient upon any point. This rule naturally came to be adopted in the Continental civil law, founded directly on the Roman law; 2 and in particular it became a part of the canon or ecclesiastical law, which for much of its material was accustomed to draw upon the Roman law. The ecclesiastical law developed the numerical principle freely, and elaborated many specific rules as to the number of witnesses necessary in various situations; against a cardinal, for example, twelve or perhaps fortyfour witnesses were required. It is enough to note that its general and fundamental rule was that single witness was in no case sufficient. In the Church's system, however, this rule received an additional sanction, over and above the mere precedent of Roman law, from the law of God as revealed in Holy Writ; for passages in the Bible, both in Old and New Testaments, were confidently appealed to as justifying and requiring this rule by Divine com

1 Digesta, xxii. 5, 12 (Ulpian: “ Ubi numerus testium non adiicitur, etiam duo sufficiunt; pluralis enim elocutio duorum numero contenta est; ") Codex, iv. 20, 4 (A. D. 283, "solam testationem prolatam, nec aliis legitimis adminiculis causa approbata, nullius esse momenti certum est; ") ib. 9, § I (A. D. 334; "Simili modo sanximus ut unius testimonium nemo judicum in quocunque causa facile pitiatur admitti. Et nunc manifeste sancimus ut unius omnino testis responsio non audiatur, etiamsi præclare curiae honore præfulgeat ").

2 This had no direct influence on our own law, and need not be further noticed. Its tenor in the 1700's may be seen in Pothier, ed. 1821, Procedure Civile, pt. 1, c. iii., and it persisted on the Continent into the 1800's.

8 Ante, 1400, Corp. Jur. Canon., Decret. Greg. lib. ii. tit. xx. de testibus, c. 23, ("licet quædam sint causæ, quæ plures quam duos exigant testes, nulla est tamen causa, quæ unius tantum testimonio, quamvis legitimo, rationabiliter terminetur;") see also, ib. c. 28, c. 4 (quoting the Bible); Decret. pars ii., causa iv., qu. ii. and iii., c. iv., § 26, reproducing Ulpian; 1713, Gibson, Codex Jur. Eccl. Angl. 1054, (“In the spiritual court, they admit no proof but by two witnesses at least; in the temporal court, one witness, in many cases, is judged sufficient; ") 1726, Ayliffe, Parergon, 541, 544, ("Though regularly single witnesses make no proof according to the civil and canon law, nor yet so much as half proof by these laws," yet there are exceptions; in criminal causes, no exception is named except for a confession;) 1738, Oughton, Ordo Judiciorum, tit. 83, p. 127 (" Jura dicunt, quod regulariter duo testes sufficiunt").

For the modern ecclesiastical law, as keeping up these rules, see Hinschius (1897), System d. katholischen Kirchenrechts, vi. p. 101, § 364.

mand; and this sanction sufficed to give to the numerical system of the ecclesiastical law an overbearing momentum and a sacred orthodoxy which must be considered in order to appreciate the force against which in due time the common law judges had to struggle.

The truth was, however, that at this time of the Papal Decretals, and long after the end of the middle ages, the rule precisely accorded with the testimonial notions of the time. It was not, in its spirit, an invention of the ecclesiastical lawyers, nor yet a mere continuance of Roman precedent; it was a natural reflection of the fixed popular probative notions of the time, notions which prevailed as well in the sturdy, self-centred island of England as on the Continent at large. The prevalence and meaning of this underlying notion must now be examined.

(2) Civilization, needless to say, almost began over again with the invasion and settlement of southern and western Europe by the Gothic hordes in the 400s and 500s. Primitive notions prevailed once more, and the slow process of development had to be repeated, repeated for the law as well as for other departments of life. Much Roman law remained in the South, and a large body of it was received in a mass in Germany in the 1500s; but this affected chiefly specific rules; the popular and general instinctive legal notions had to grow once more out of primitive into advanced forms. Now one of the universal and marked primitive notions is that of the oath as a formal act, mechanically and ipso facto efficacious (like the ordeal and the trial by battle), and quantitative in its nature. This notion is merely one particular phase of the entire system of formalism inherent in the stage of intellectual development at which our Germanic ancestors were at that epoch. It is a matter of the whole spirit of the times, not of a particular or local belief; and since the history with which we are now concerned is that of the growth and change of a radical and epochal conception, not easy to reproduce in our modern imaginations, it may be worth while (for obtaining a starting point) to

1 Deut. 17, 6: [“ The murderer shall be put to death; ] but at the mouth of one witness [only] he shall not be put to death;' 19, 15: "For any iniquity. . . at the mouth of two witnesses, or at the mouth of three witnesses, shall the matter be established;" Numb. 35, 30 (like Deut. 17, 6); Matt. 18, 16: ["If thy brother trespass against thee, and reject thy complaint,] then take with thee one or two more, that in the mouth of two or three witnesses every word may be established; " II. Cor. 13, I (similar); I. Tim. 5, 19; Hebr. 10, 28 (allusions to the foregoing ideas); John 8, 17: "It is also written in your law that the testimony of two men is true."