Of course, in saying that this peculiarity has not been considered, I am not suggesting that it has not been noticed, or that its cause is in any way doubtful or unknown. We know that the earth, besides whirling once a day on its axis, and rushing on its mighty orbit around the sun (spanning some 184,000,000 of miles) reels like a gigantic top, with a motion so slow that 25,868 years are required for a single circuit of the swaying axis around an imaginary line upright to the plane in which the earth travels. And we know that in consequence of this reeling motion the points of the heavens opposite the earth's poles necessarily change. So that the southern pole, now eccentrically placed amid the region where there were no constellations in old times, was once differently situated. But the circumstance which seems to have been overlooked is this, that by calculating backwards to the time when the southern pole was in the centre of that vacant region, we have a much better chance of finding the date (let us rather say the century) when the older constellations were formed, than by any other process. We may be sure not to be led very far astray; for we are not guided by one constellation but by several, whereas all the other indications which have been followed depend on the supposed ancient position of single constellations. And then most of the other indications are such as might very well have belonged to periods following long after the invention of the constellations themselves. An astronomer might have ascertained, for instance, that the sun in spring was in some particular part of the Ram or of the Fishes, and later a poet like Aratus might describe that relation (erroneously for his own epoch) as characteristic of one or other constellation; but who is to assure us that the astronomer who noted the relation correctly may not have made his observation many hundreds of years after those constellations were invented? Whereas, there was one period, and only one period, when the most southernmost of the old constellations could have marked the limits of the region of sky visible from some northern region. Thus, too, may we form some idea of the latitude in which the first observers lived. For in high latitudes the southernmost of the old constellations would not have been visible at all, and in latitudes much lower than a certain latitude, presently to be noted, these constellations would have ridden high above the southern horizon, other star-groups showing below them which were not included among the old constellations.