whose intro-moving winds, bearing the evaporations of the Asiatic seas and oceans, feed it with meteoric fuel for six months in the year, and whose periphery may be regarded as embracing nearly one-third of the entire Eastern Hemisphere. Analogy, therefore, warrants the idea of a great cyclone. But, apart from all this, actual observations in different parts of the globe prove the frequency of storms of enormous magnitude. Thus, in the celebrated Gulf-Stream storm of 1839, as Sir David Brewster long ago pointed out, several stanch merchant-men were foundering off the coast of Georgia, near Savannah, in the very heart of the gale, at the same hour that the winds in its north-west quadrant were taking the roofs off houses in New York and Boston, more than 800 miles distant—clearly revealing a cyclone whose formation was symmetrical, and whose diameter must have been nearly 1,300 miles. But, not to go back to old data, the West-Indian storm of the 18th of August, 1871, before its centre had moved north of Florida, had begun to draw upon the regions of high barometer in the northern States, had exerted its influence as far north as New London, Connecticut, and gave us the northeasterly cyclonic winds in the northwest quadrant of the whirl, on the entire Atlantic coast. The more furious cyclone of the 24th of August, discovered to be then south-east of Florida, and telegraphically foreannounced as likely to endanger the coasts of the Southern States in less than forty-eight hours, appeared on the 26th in full force in Northern Florida, but not until some eight or ten hours after it had set the atmosphere all around it (as far north as Boston) in cyclonic motion, and had caused the storm-cloud to spread itself over the entire region of the United States on the eastern slopes of the Alleghanies, and as far westward as Knoxville, Tennessee. It is no uncommon thing, as Redfield, Espy, Henry, Loomis, and others, long ago showed, for an area of depression on the upper lakes to make itself simultaneously felt as far south as the Gulf of Mexico, and as far east as New England.
If it fell within the scope of the present design in this paper to consider the final cause of storms, it would be easy to show that, unless the law of storms ordained a large area, and a far extended path for the meteor, in some degree commensurate with the area of our immense continent, the meteor could not fulfil its office in the terrestrial economy—an office which, apparently, imposes upon it the task of gathering to its centre, through the agency of its intro-moving winds, the idle and inappreciable moisture scattered over the surface of the earth, condensing it into rain and snow, and diffusing it, in these forms, over immense districts of country.
It is of incalculable importance to observe, and carefully digest the fact, that, when a storm-centre or area of low barometer is once formed, it is the nucleus for a vast aggregation and marshalling of meteoric forces. No matter how small at first, under favorable atmospheric conditions, the courant ascendant is formed, condensa-
