When a thunder storm, proceeding in either a western or eastern direction, as the case may be, happens to strike a large water-course running either north or south, and when also there happens to be a large branch emptying into the stream, within a few miles either above or below the point where the storm approaches it, I have uniformly observed the storm to cross the large stream at the point where the large branch unites with it, and ascend the branch. Where there are two large tributaries about equi-distant from the point of approach, the storm frequently divides and follows each of them. The reason why it should be so, this is not the place to discuss; but the Wisdom and Goodness which so ordered it, are too apparent to every rational mind to be overlooked. It may be asked if the difference in latitude and elevation between the Ohio and lake regions of country, does not produce a great difference in the climates of those respective regions? These causes certainly produce some difference, but not all. It is my object to establish facts, rather than any favourite theory. The difference of latitude between the Ohio river at the mouth of the Scioto, and lake Erie at the mouth of the Maume or Sandusky, is nearly three degrees, and the difference of elevation above the sea is trifling, if any. From the mouth of the Scioto to Columbus, about 90 miles in a direct line, the water, where there is what is commonly called a _ripple_, runs briskly, and these ripples happen, perhaps, one to a mile; but they are in a sandstone region, and the fall of course is trifling. Let us suppose then, that the river Scioto descends one hundred feet from the mouth of the Whetstone, which empties into that river at Columbus, to the Ohio, and that the Whetstone which runs through a limestone formation, descends another hundred feet, which would make Upper Sandusky two hundred feet higher than the Ohio river. From this highest ground between the Ohio and the lake, it is a well-known fact, that the land descends towards the north much more in a given distance, than it does towards the south, and the distance is not half as far. The Maume and other streams putting into the lake, are full of rapids. Admitting for argument's sake, that the Sandusky or Maume descend only 100 feet, then the surface of the lake is 100 feet higher than the Ohio river. Would three degrees of latitude, and 100 feet greater elevation produce three weeks difference in the seasons? Is there that difference between Baltimore in Maryland, and Wilkesbarre in Pennsylvania? Is there that difference between New-York and Fort Edward on the Hudson? It is believed that there is not one half that difference. I have referred but little to thermometers, because they are kept in so many different situations by their owners, that I have known no less than 8 degrees of difference between several of them kept in one town, within almost a stone's throw of each other, at one and the same moment of time.