Thucydides is our noble legislator; he marked the admiration that met Herodotus and gave the Muses' names to his nine books; and thereupon he drew the line which parts a good historian from a bad: our work is to be a possession for ever, not a bid for present reputation; we are not to seize upon the sensational, but bequeath the truth to them that come after; he applies the test of use, and defines the end which a wise historian will set before himself: it is that, should history ever repeat itself, the records of the past may give present guidance.

Such are to be my historian's principles. As for diction and style, he is not to set about his work armed to the teeth from the rhetorician's arsenal of impetuosity and incisiveness, rolling periods, close-packed arguments, and the rest; for him a serener mood. His matter should be homogeneous and compact, his vocabulary fit to be understanded of the people, for the clearest possible setting forth of his subject.

For to those marks which we set up for the historic spirit—frankness and truth—corresponds one at which the historic style should first of all aim, namely, a lucidity which leaves nothing obscure, impartially avoiding abstruse out-of-the-way expressions, and the illiberal jargon of the market; we wish the vulgar to comprehend, the cultivated to commend us. Ornament should be unobtrusive, and never smack of elaboration, if it is not to remind us of over-seasoned dishes.

The historian's spirit should not be without a touch of the poetical; it needs, like poetry, to employ impressive and exalted tones, especially when it finds itself in the midst of battle array and conflicts by land or sea; it is then that the poetic gale must blow to speed the vessel on, and help her ride the waves in majesty. But the diction is to be content with terra firma, rising a little to assimilate itself to the beauty and grandeur of the subject, but never startling the hearer, nor forgetting a due restraint; there is great risk at such times of its running wild and falling into poetic frenzy; and then it is that writers should hold themselves in with bit and bridle; with them as with horses an uncontrollable temper means disaster. At these times it is best for the spirit to go a-horseback, and the expression to run beside on foot, holding on to the saddle so as not to be outstripped.

As to the marshalling of your words, a moderate compromise is desirable between the harshness which results from separating what belongs together, and the jingling concatenations—one may almost call them— which are so common; one extreme is a definite vice, and the other repellent.

Facts are not to be collected at haphazard, but with careful, laborious, repeated investigation; when possible, a man should have been present and seen for himself; failing that, he should prefer the disinterested account, selecting the informants least likely to diminish or magnify from partiality. And here comes the occasion for exercising the judgement in weighing probabilities.

The material once complete, or nearly so, an abstract should be made of it, and a rough draught of the whole work put down, not yet distributed into its parts; the detailed arrangement should then be introduced, after which adornment may be added, the diction receive its colour, the phrasing and rhythm be perfected.

The historian's position should now be precisely that of Zeus in Homer, surveying now the Mysians', now the Thracian horsemen's land. Even so he will survey now his own party (telling us what we looked like to him from his post of vantage), now the Persians, and yet again both at once, if they come to blows. And when they are face to face, his eyes are not to be on one division, nor yet on one man, mounted or afoot—unless it be a Brasidas leading the forlorn hope, or a Demosthenes repelling it; his attention should be for the generals first of all; their exhortations should be recorded, the dispositions they make, and the motives and plans that prompted them. When the engagement has begun, he should give us a bird's-eye view of it, show the scales oscillating, and accompany pursuers and pursued alike.

All this, however, with moderation; a subject is not to be ridden to death; no neglect of proportion, no childish engrossment, but easy transitions. He should call a halt here, while he crosses over to another set of operations which demands attention; that settled up, he can return to the first set, now ripe for him; he must pass swiftly to each in turn, keeping his different lines of advance as nearly as possible level, fly from Armenia to Media, thence swoop straight upon Iberia, and then take wing for Italy, everywhere present at the nick of time.

He has to make of his brain a mirror, unclouded, bright, and true of surface; then he will reflect events as they presented themselves to him, neither distorted, discoloured, nor variable. Historians are not writing fancy school essays; what they have to say is before them, and will get itself said somehow, being solid fact; their task is to arrange and put it into words; they have not to consider what to say, but how to say it. The historian, we may say, should be like Phidias, Praxiteles, Alcamenes, or any great sculptor. They similarly did not create the gold, silver, ivory, or other material they used; it was ready to their hands, provided by Athens, Elis, or Argos; they only made the model, sawed, polished, cemented, proportioned the ivory, and plated it with gold; that was what their art consisted in—the right arrangement of their material. The historian's business is similar—to superinduce upon events the charm of order, and set them forth in the most lucid fashion he can manage. When subsequently a hearer feels as though he were looking at what is being told him, and expresses his approval, then our historical Phidias's work has reached perfection, and received its appropriate reward.