18G TRANS. OF THE ACAD. OF SCIENCE. 



tions have been uninterruptedly kept up with the best instruments this 

 country can afford, viz., those adopted by the Smithsonian Institution, and 

 constructed by Mr. James Green of New York. 



Three of these stations, Superior City, Ontonagon, and Marquette, are 

 situated on Lake Superior; two, Milwaukie and Grand Haven, on Lake 

 Michigan; two, Thunder Bav and Ottawa Point, on Lake Huron; one, 

 Detroit, on Detroit River; three, Monroe Piers, Cleveland, and Buffalo, on 

 Lake E r ie; and three, Fort Niagara, Charlotte, and Sackett's Harbor, on 

 Lake Ontario. 



The observations here recorded with such excellent instruments, if con- 

 tinued in a sufficiently lengthy series, will constitute a mass of trustwor- 

 thy data, upon which, not only the present question can be satisfactorily 

 settled, but" a great many others of much magnitude in scientific aspects, 

 as well as practical utility, may be scrutinized in a spirit presaging per- 

 fect success in the establUhment of the most important truths connected 

 with the subject, in its widest range. Not only are these valuable obser- 

 vations kept up, but all the improvements of moilern science are added to 

 the efficiency of the instruments, by that accomplished officer the succes- 

 sor to Cap*-. Meade in the Superintendency of the Lake Survey, Col. J. D. 

 Graham, of the same Corps. In addition to the fact established by the 

 late Prof. Espy, Prof. Loomis, and others, in their studies of the great 

 storms of the United States, that the minimum wave of the barometer 

 which usually precedes and accompanies them — especially those of the 

 winter, which are known to pervade that portion of the United States 

 embraced between the Rocky Mountains and the eastern coasts of the At- 

 lantic Ocan — proceeds in an easterly direction ; it is further developed by 

 the comparisons of the barometer at Superior City with those at Thunder 

 Bay and Sackett's Harbor, that nearly all the motions of the barometer 

 are re-produced, or succeed each other in a regular order of time within 

 the limits of about 7 to 48 hours between those places, always beginning 

 at the western station, showing themselves next at Thunder Bay, thence in 

 the same apparent order, on to Sackett's Harbor. For want of a more ex- 

 pressive term we call this a wave of the barometer. Not only this wave, 

 whether it be a minimum or a maximum, but also the intermediate motions 

 of the barometer, appear to observe the same rule of progression from 

 west to east. 



Out of one hundred and fifty great culminations of the barometer, when 

 it rises considerably above the monthly mean, from the sudden fall of 

 which rain or wind storms may be predicted, besides many others near or 

 but slightly above the mean, there are in two years but three exceptions 

 where the changes take place, first at Sackett's Haibor, and the wave ap- 

 pears to travel from east to west. There are, moreover, in this time, two 

 other exceptions which show the motions at all the places simultaneously. 

 Here, it appears, that whatever cause influences the movements of the 

 barometer, it is so pervading and extensive as to reach from Superior 

 City to Sackett's Harbor, a distance of some 16° of longitude. 



Before the regular establishment of the Lake Survey meteorological 

 stations, a series of observations were kept at Forrestville and Thunder 

 Bay, two contiguous stations, commencing June, 1858, and ending May, 

 1860, which, by comparison with simultaneous observations at Toronto 

 and Providence, kept at the first place by Prof. Kingston, and at the lat- 

 ter by Prof. Caswell, shows almost exactly the same motion of the wave 

 of the barometer from west to east, as those developed by the regular ob- 

 servations of the Survey for the two succeeding years. In these observa- 

 tions, the exceptions of the wave moving westward are also three, in two 

 years ; but there appears in this time to have been no pervading wave, 

 reaching from one end of the line of stations to the other, as shown on 

 the curves of the barometer from 1860 to 1862. 



If, instead of the tri-daily, hourly observations could be substituted at 

 the extreme and middle Nations, much fairer results, both with regard to 

 the range of the barometer and the more exact time and rapidity of 



