But the merits, at first, outweigh the faults. Until he runs up against the beginnings of Rome—with all the irritation which the mere name of Rome provokes in him—he keeps his head and writes excellently. The thing is well balanced and of real value. The exceptions which must be made, even to the part before Greece and Rome, to this praise do not colour his story of early record as a whole. He gives a rather imaginary account of the beginnings of agriculture, but much of it is more probable than improbable; and he modifies it well enough by conditional adverbs proper to our necessarily speculative attitude towards these early and uncertain things. Now and then, in a sentence or two, he is unwise enough to abandon this conditional attitude and bolts away again into fiction: for instance, he tells us that when man settled down to agriculture, the Red Sea was still connected with the Mediterranean (p. 91). We do not know that; it is mere guesswork, and ought not to be put up as historical fact. On the other hand, an immediately following remark, that “the Persian Gulf then extended much further northwards than it does now,” is real history; for there is sufficient proof of that. Again (on p. 92), he puts down as historical fact the invariable conquest of settled populations by barbaric and nomadic populations outside. He treats it as a necessarily recurrent phenomenon and as the only process. Of course, we all know from History that in plain recorded fact the converse is just as common, and far more lasting in its effects. Subjugation of the barbarian by the civilized man is very much more the rule of recorded History than its opposite. And as for agricultural work being regarded as the lot of an inferior, that is a false generalization from our own suburban conditions. On the contrary, the tradition of the Mediterranean peoples, as of the Chinese, is just the other way. Agricultural work was the basis of their society and was clothed with moral dignity. The account of early Mesopotamian civilization (including in that term the earliest culture before the junction of the Euphrates and Tigris systems) is excellent, and the subsequent short section on the beginnings of Egypt is equally good—with a paragraph on the causes of the difference of record (on p. 98) between the Nile and the Euphrates Valleys particularly well put. In order to keep the survey parallel and marching on one front, there follow a few lines on the early civilization of India and on the early history of China; in which, by the way, Mr. Wells rightly rejects the somewhat shadowy hypothesis put forward of late years, which gives a Mesopotamian origin to the Chinese culture.