208 NINTH ANNUAL REPORT OP 



.ature correspond with 1-lOlh inch of mercury in the barometer, I find 

 some very interesting results are brought out. In all the States south 

 and west of the Alleghany range, the parallelism between the changes 

 in the barometer and the changes in temperature are much greater than 

 I anticipated. See figures Nos. 1, 2 and 3. 



Some remarkable exceptions, however, occur, in which a rise of tem- 

 perature is accompanied with a rise of the barometer. These excep- 

 tions are found to the east of the Alleghanies and the mountains of New 

 England, where the curves of temperature and pressure of the air en- 

 tirely lose the regularity which they possess in the southern and western 

 States. The changes of the barometer before storms in the northeast- 

 ern States appear to be very peculiar, at least I am acquainted with 

 nothing like them in Europe. When the mercury rises a little above 

 30 inches in Great Britain, the index of the barometer points to "set fair," 

 and an eastern storm rarely comes on before the mercury begins to 

 sink. When the mercury is above 30 inches in Maine, during winter 

 or autumn, a storm is often at hand, and the barometer does not fall be- 

 fore the weather gets colder and drier. An examination of the weather, 

 from the 7th to 16th November last, (see figure 4,) apparently shows 

 that the indications of the barometer, as a weather prophet, must be 

 interpreted by a contrar}^ rule to that which is observed in the western 

 States and in England. The probable cause of this will be afterwards 

 adverted to. I shall now proceed to give you an account of a general 

 storm, which will serve to illustrate the principles I have stated : 



The weather on the 12th November, 1854, presented a curious pic- 

 ture of extremes over the United States. The northwest wind was 

 advancing like an extended wall, from Iowa to Texas, and clearing 

 the whole country of its moisture. By looking at the following table 

 of temperature for that day, we shall see the progress which it had 

 made at the time mentioned: 



Iowa, Poultney 11° 7, a. m. 



Wisconsin, Madison 19<^. " 



Missouri, St. Louis 26^ " 



Mississippi, Oxford 34°. 9, p. m. 



Texas, New Wied 36° " 



Oq the morningof the 12th the first killing frost was felt in Louisiana, 

 and over Michigan, the northern parts of Illinois, Indiana, and Ohio, the 

 ground was covered for the first time with snow. In these States the 

 northwest wind was flowing beneath the colder edges of the southwest 

 stream from the Gulf, forcing it to ascend and precipitate part of its 

 moisture in the form of snow. As the under current rose it was caught 

 by the upper current, and, as in the case of thunder storms, it was 

 swept off towards the east. This at least seemed to be the process at 

 Indianapolis, where, on the 12th, four inches of dry snow fell, while the 

 lower wind from the north and northwest was overlaid by a current 

 bearing clouds from the southwest. On the morning of the 13th, at the 

 same place, the wind set in from the west. 



On looking at the meteorological chart for the morning of the 12th, we 

 perceive at a glance that the temperature gradually rises from Iowa to 

 Maine, from Missouri to Savannah, and iirom Texas to Florida. The 



