We therefore decline to enter into disquisitions, interminable and surely disappointing, upon the rivers Pishon and Gihon, and the countries of Havilah and Cush. Etymological similarities afford no safe ground for conclusions ; for many names of close resemblance are to be found in the Asiatic languages, but of which the natural history and collateral circumstances are incompatible with other parts of this (as we think) antediluvian fragment of topography. Also Gihon certainly, and probably Pishon, were used in the ancient Oriental lan guages as appellatives, separate or prefixed, signi fying a stream in general ; as the old British Avon, which has the same meaning, has become the proper name of several rivers in England, Wales, and Scotland.
III. We-venture to give a summary of this de scription. It was a tract of country, the finest ima ginable, lying probably between the 33d and the 37th degree of N. latitude, of such moderate ele vation, and so adjusted, with respect to mountain ranges and watersheds and forests, as to preserve the most agreeable and salubrious conditions of temperature and all atmospheric changes. Its sur face must therefore have been constantly diversified by bill and plain. From its hill-sides, between the croppings out of their strata, springs trickled out, whose streamlets, joining in their courses, formed at the bottom small rivers, which again receiving other streams (which had in the same way flowed down from the higher grounds), became, in the bottom of every valley, a more considerable river. These valleys inosculated, as must consequently their contained streams; wider valleys or larger plains appeared; the river of each united itself with that of its next neighbour; others contributed their waters as the augmenting stream proceeded; and finally it quitted the land of Eden, to continue its course to some sea, or to lose its waters by the evaporation of the atmosphere or the absorption of the sandy desert. In the finest part of this land of Eden, the Creator had formed an enclosure, pro bably by rocks and forests and rivers, and had filled it with every product of nature conducive to use and happiness. Due moisture, of both the ground and the air, was preserved by the streamlets from the nearest hills, and the rivulets from the more distant; and such streamlets and rivulets, collected according to the levels of the surrounding country (` it proceeded from Eden'), flowed off afterwards in four larger streams, each of which thus became the source of a great river.
This metaphrase deviates from what is commonly thought to be the meaning of the original, but not, we think, from its true signification and intention.
1. It is a metonymy occurring probably, though not very frequently, in all languages, that a col lective noun is sometimes used when the idea is compound and distributive. The usage is recog nised in the Hebrew language, by Gesenius in his Lehrgebliude, p.525; Ewald, Gramm., sec. 346 ; and Nordheimer, Gramm. sec. 738-750. This
kind of synthesis would be likely to find place in a primitive and consequently very simple language. The multitude of droppings and tricklings, rills and streamlets, having one beneficial design, and ever tending to confluence, would, in the mind of a primeval writer, readily coalesce into a singular term, a river. We have an appropriate example in Ps. lxv. to, where the aggregate of showers is called the river of God, full of water.' The principle applies equally to ha and Nu. It is therefore no unwarrantable liberty to understand by the river,' a number of rills and rivulets dis persed throughout the ground, and flowing into one channel about the issue into the external country. If the water entered the garden as a river properly- that is, in one body—it could not water the garden' without artificial appliances ; and it would have divided the garden, making one part inaccessible from the other, without a boat or a bridge.
2. That a river should be divided into four heads,' or sources of new rivers, is naturally impos sible. If to a running stream, small or large, two or more channels be presented, it will not divide itself distributively, but will pour its whole mass of water into the deepest channel : it will ever seek the lowest bottom. We must therefore understand the passage as saying that, from four different collec tions of rills, which had flowed down different de clivities in the same neighbourhood, the sources were formed of four rivers, which in their progress became great and celebrated. To controvert this reasoning it would not be sufficient to adduce the division of a great river into branches as it ap proaches the sea, and meets an extensive swamp or flat shore, as in the deltas of the Rhine (forming, with many inferior streams, the Leek and the Waal), the Po, the Nile, the Ganges, and many others. The soft and almost horizontal level causes the water to cease flowing, or nearly so, and the vast extent of mud or sand permits branches of the stream to take place when some small change of the surface gives occasion. But the rivers of Paradise must have been in high ground, and have had a considerable fall. It is possible, indeed, that rocky obstacles might exist, connected backwards with a mountainous country, presenting their heads against the stream, and thus separating it, as islets are formed in the higher course of the Rhine. But the conditions necessary to derive four great rivers out of one, in this way, are scarcely conceivable as occurring in one place. The origin of two or more rivers from different fountains in the same locality of high ground, hut on different levels, and then pursuing different courses, is not an unexampled phenomenon. The Rhine and the Rhone rise but about eight English miles from each other ; and, which applies to the case directly before us, the sources of the Euphrates and the Tigris, on the eastern frontier of Armenia, so far as they can be followed up, are only fifteen miles apart.