THE SMITHSONIAN INSTITUTION. 201 



Stratum, we must in general look for a local cause of propulsion, on the 

 same principle that the motion of the air at the surface of the earth, 

 during the day in summer, with clear and dry weather, is constantly 

 varied by the heat of the sun at every spot over which the breeze 

 passes. As soon as the warm southerly wind begins to blow as a gen- 

 eral surface current over the United States, the conditions necessary to the 

 production of a storm already exist. The southerly wind, often 

 veering about very irregularly, will then frequently continue in the same 

 direction, until the whole country from the Gulf of Mexico to Canada is 

 covered with a comparatively hot and moist stratum of air. Under 

 these circumstances a crisis must ensue, when the upper current from 

 the Rocky Mountains resumes its sway at the surface of the earth. 

 In winter it often sweeps the stratum of moist air from over the 

 United States into the Atlantic, and completely changes the character 

 of the weather in a very short time. In summer the changes are gen- 

 erally much less marked ; for Professor Coffin's researches show that 

 the northerly winds then lose their power. At ihis season the process 

 usually takes place in a gradual manner. Thunder clouds, drifting 

 in the upper current, as described by Professor Dwight, shower 

 down their watery contents over the land, and thus enable Indian 

 corn to be successfully cultivated over a surface of country vastly 

 greater than in any other part of the world. How these effects are 

 brought about, involves the whole question of the action of storms east 

 of the Rocky Mountains. We shall leave, however, the further discus- 

 sion of this for the present, and conclude this lecture by one example, 

 showing the scale upon which thunder storms are sometimes developed 

 much about the same time over a vast area of country. I am only 

 sorry that my time has not permitted me to analyze this storm in 

 all its details. 



During the first days of last September the wind was mostly from the 

 south ; the weather became excessively hot and oppressive ; the news- 

 papers in all parts of the country were recording the high temperatures ; 

 when, on the 6th, thunder storms took place nearly simultaneously in 

 Iowa, Illinois, Indiana, Ohio, Pennsylvania, New York, East and West 

 Canada, and the New England States. Large quantities of rain fell in 

 various parts of the country, as the storms were in several places 

 somewhat locally developed. At some points the northwest upper 

 current reached the surface of the earth for a time, the southwest wind 

 again blowing as before, until a general change of wind to the north- 

 west prevailed, and caused a great fall in the temperature. At Sara- 

 toga the thermometer stood at 96° in the shade on the afternoon of the 

 6th, and at 46^ on the morning of the 9th at Rochester. It is a fact 

 worthy of attention that a severe storm, amounting to a hurricane, 

 swept the southeastern coast of the United States just about the time 

 that this great change was taking place in the north and west. It is 

 certainly well worthy the investigation of American meteorologists to 

 ascertain whether any connexion exists between the weather in the 

 northwestern States and the hurricanes of the West Indian islands, 

 for in this instance the coincidence of phenomena was quite remark- 

 able. As will be seen from our next lecture, great changes occurring 

 in the northwest are rapidly propagated to the southeast in the case of 

 winter storms. 



