200 Dr. Bostoclt's Observations on [March, 



three lines traced xrpon the plan, it will be observed that the 

 Liverpool barometer is generally the lowest, while that in 

 London is the highest ; but these heights are not proportional 

 to the actual situation of the instruments ; the London and 

 Liverpool barometers being nearly on a level, each about 80 feet 

 above low water in the Thames and Mersey respectively, while 

 that at Lyme was about 100 feet higher. But what is more 

 remarkable, the proportional heights were perpetually varying, 

 so that neither the time nor the degree of the rise and fall cor- 

 respond. It happens, on more than one occasion, that, upon any 

 considerable rise or fall of the mercury, the maximum or minimum 

 occurred on different days ; and in some instances we observe 

 that the mercury was even rising in one of the barometers while 

 it has been falling in another. Upon the whole, the London and 

 Liverpool barometers observed a more uniform course with 

 respect to each other than the one at Lyme, although they 

 differed most in their absolute height. 



It is easy to conceive that various local causes may affect the 

 atmosphere, so as to render it heavier in one situation than in 

 another, although not very distant from the former : certain 

 winds blowing over a large extent of the ocean, or even over 

 particular soils, the vicinity of mountains, the position of rivers 

 or deep valleys, perhaps even of large cities, may possibly affect 

 the weight of the air. But before we proceed to any investiga- 

 tions of this nature, it will be necessary to inquire how far the 

 instrument itself, by which these changes are ascertained, is to 

 be implicitly relied on. The principle upon which the baro- 

 meter acts is so simple, that when the apparatus is in a perfect 

 state, it would be liable to no uncertainty or irregularity ; but 

 although the principle of the instrument be correct, its construc- 

 tion is very frequently defective ; and indeed its very nature is 

 such as to render it peculiarly liable to become deranged from 

 various causes, even when it had been accurately constructed 

 in the first instance. 



Before, therefore, I proceeded any further in the investigation, 

 or attempted to form any theory to account for the diversity of 

 the barometers at the three stations of London, Liverpool, and 

 Lyme, I thought it necessary to institute a comparison between 

 two or more barometers that were placed nearer together, and 

 where the differences, if they existed, must be ascribed to some 

 inaccuracy in the instrument, rather than to any actual variation in 

 the weight of the atmosphere. For this purpose I had recourse 

 to the meteorological diaries that were kept at that time by Mr. 

 Howard and Mr. Forster, the one at Plaistow, and the other at 

 Clapton, at a distance of not more than four or five miles from 

 each other. These gentlemen are well known for the attention 

 which they have paid to all points connected with the weather ; 

 and with respect as well to science as to accuracy, may be 

 regarded as affording us the means of establishing the most 



