Having thus taken a brief view of the different clouds, let us return to the inquiry, from what ocean, and by what machinery, _our_ "rivers return." Not wholly or mainly from the North Atlantic, although it lies adjacent to us, and they often _seem_ to do so; for, first, all storms, showers, and clouds, which furnish, _independently_, any appreciable quantity of rain to the United States, and even adjacent to the Atlantic, or indeed to the Atlantic itself, come from a westerly point, and pass to the eastward. _This is a general, uniform, and invariable law, although there is in different places, and in the same place at different times, some variation in their direction; ranging in storms from W. by S. to S. S. W., and in showers between S. W. and N. W., to the opposite easterly points of the compass; the most general direction, east of the Alleghanies, being from W. S. W. to E. N. E._ But do we not see, you inquire--at least those of us who live east of the Alleghanies--that when it rains, the wind is from the eastward; and that the _clouds_ follow the wind from the east to the west? You do indeed, generally, in all considerable storms, observe that the wind blows from some easterly point, and that _seeming_ clouds are blown by it to the westward; but what you see, and call clouds, are not the clouds which furnish the rain. Far above the seeming clouds you notice, directly over your head when it rains or snows, are the rain or snow clouds, dense and dark, passing to the eastward, how strong soever the wind may blow from the quarter to which they tend, or any other quarter, between you and them. What you see below them are _scud_. So the sailors call them, and so I have termed them. It is a "dictionary name," and a good one, expressive of a distinction between them and _clouds_. They are thin, and the sun shines through them, although with some difficulty, when the rain clouds above are absent or broken. _This east wind and the scud are not the storm, or essential parts of it._ Storms occasionally exist, particularly in April, without either. They are but _incidents_, _useful_, but not _necessary incidents_, as all surface winds are. If you could see a section of the storm, you would see the rain cloud above, moving to the east, and the scud beneath running to the west, as indicated by the arrows in the cut on page 40. Opportunities frequently occur when these appearances may be seen. Storms are sometimes very long, a thousand miles, perhaps, from W. S. W. to E. N. E., and not more than one to three hundred miles wide from S. E. to N. W., and their sides, particularly the northern ones, regular, and without extensive partial condensation. Then the storm cloud above, moving to the eastward, and the scud running under to the westward, may be seen as in the cut.