which he raised but was not permitted to use the long row of well-filled notebooks, his notes of references to learned periodicals and even his cuttings from the daily press whence he thought he might throw some light on Roman times, or perhaps only obtain a telling comparison from some modern event or institution. If one may venture a criticism, he carried this virtue too far, and the unusual range of his interests sometimes induced him to spend his time on the pursuit of some minor point, forgetting how great a task he had still to carry out and how many bye-paths of ancient history are likely in the present state of the evidence to prove blind-alleys.

Such a disposition is free at all events from the temptation to rash judgments, and the very amount of his material made Arnold doubly careful in drawing inferences. On the first page of all the chapters now published the margins are crowded in their original form with memoranda queries whether he had overemphasised this point or understated that, a note that he must read or re-read some German book or article before passing for press, or a vigorous criticism of his own style or arrangement in some particular section. Such features do not indicate a temperament likely to facilitate an author's progress, but what is done is well done, and their sobriety stands in pleasant contrast with the ill-informed dogmatism of much that has been written on Roman History in modern times.

The fact that we have only a fragment before us, a fragment which itself was considered by the author to need much further labour, makes it specially hard to decide what was Arnold's conception of history. But even from a fragment some estimate may be attempted if we remember how conjectural such an estimate must be.

J

In the view of history as a theatre for the unfolding of great personalities he had comparatively little interest. This may have been in part owing to the period with which he dealt. The golden age of Roman character which furnished such noble studies to Livy and Plutarch was past. The age of Augustus was too placid for the play of the heroic virtues. The outstanding figures of the early Empire-for we may pass over the futile aristocrats who posed as the inheritors of the Republican tradition-were either persons ruined by absolute power and unlimited license or vigorous, hard-working statesmen with the prosaic virtues of the successful business man. On the former class-a class who have formed the theme of much eloquence ancient and modern-Arnold spends little time, passing over the vices and follies of Julia as briefly as his subject will allow. His only fulllength portraits are Augustus and Agrippa, the one a strange mixture of selfishness and patriotism, ruthlessly sacrificing the lives of his enemies, his own ease, and the happiness of his family to the new composite deityRoma et Augustus; the other the laborious, able administrator, resigned to the knowledge that he could never aspire to the first place.

History in Arnold's view dealt more with communities than with individuals. Augustus' character was of importance not so much for its psychological interest as for its being a factor in reshaping the Roman Constitution and in modifying the relations of Rome and her subjects, that is, in its constitutional and imperial consequences. Two of the chapters are devoted to the Constitution; that he did not dwell longer on it was doubtless due to the fact that the field had been thoroughly worked over by his predecessors, and while here and there minor points remained in dispute it was only necessary to

restate in brief space the results already ascertained. Moreover the exact division of power between Senate and Emperor was not of prime importance. From the first the Emperor was the predominant partner, and his share of power was always becoming larger.

The real point of importance for the historian of that period is not the relation of the two partners in the government, but the relations between governors and governed. The artificial arrangements of the Dyarchy had no influence on subsequent ages; it is superfluous to explain the influence of the Roman Empire for the last two thousand years.

Arnold had grasped this fact. Even in his two chapters on the Constitution a part is devoted to general provincial administration, and in what follows he deals mainly with the several provinces of the Empire. He made it his task to show the condition of these lands before the coming of the Romans and the conditions resulting from the contact of the two diverse elements, and to point out how Roman influence was deeper in one place and weaker in another, and what was the limit at which it ceased to exist.

There was an additional attraction to Arnold in this aspect of his subject. Through his work as a journalist, perhaps also by a natural bent, he was deeply interested in all imperial questions. The Roman Empire was the first great Imperial experiment which rose above the methods of brute force or mere well-devised bureaucracy. Rome made a genuine effort to unite Liberty and Empire, and though she ultimately failed she offered, if not political lessons, which it is always hard to deduce correctly, at least a highly interesting analogy to similar modern experiments. In particular the English historian is irresistibly reminded of the British Empire, and especially of its great Indian dependency. To this Arnold's thoughts often turned.

The abolition of the darker side of the religion of the Druids recalled the suppression of Suttee and Juggernaut. The buffer state of Siam reminded him of the similar defence erected by Rome on the Euphrates. The Client Princes of the Roman Empire came for him under the same category as the protected rulers of India. In the collections which he made for his history there is much that deals with modern imperial topics-an article in the Asiatic Quarterly Review on Native India by Sir Lepel Griffin, a speech by Sir Henry Fowler on Indian Government, or an article on Tonkin in the Journal des Débats. It is possible that the search for analogies of which he was so fond may have taken him too far and have caused him to forget, as most historians except Finlay have forgotten, the terrible price which was paid first by the conquered and then by the conquerors for the establishment of the Roman Empire. Its excellent organisation need not blind us to the fact that after all the Romans were still near the period when a war of conquest was waged with the most frankly materialistic aims.

There were other reasons which attracted Arnold to the subject. In the preceding memoir it has already been shown how fond he was of geography and travel, how interested in foreigners, how anxious that a better understanding should prevail between them and his fellowcountrymen. All this gave additional zest to his study of an empire which had spread over the lands now occupied by great modern nations. He was a diligent visitor of Roman sites, and the knowledge thus acquired lends picturesque touches to his narrative, as when he mentions the black walls and red roofs of Nyon, or that "miracle of grace and strength" the Pont du Gard. And it was not only such monuments as the Arch at Igel or the great rampart stretching from

Rhine to Danube which were memorials to him of Roman days, but also some touch of custom or language, the black dress of a Spaniard or the name of a French Bishopric, while the Provençal tongue recalled to him how in that corner of France, that Italy in Gaul, Rome had established herself longest and most securely.

His survey of the Provinces was still unfinished when he died. Such important parts as Roman Africa and Roman Britain were wanting. If we judge by mere bulk, he had accomplished little. But, apart from the quality of his writing, which it has been the object of the preceding paragraphs to explain, it is well to remember that the influence of a scholar's life may not be measured by his published work. It was not only that he freely communicated his knowledge and advice or lent his books freely to those who asked. His learning was an instrument of culture for character as well as for intellect, and such culture may exercise as deep and permanent an influence as printed books.