366 On Storms and JWeteoroIogical Observations. 



The mean temperature of any place is so very readily ascertained 

 and with so much ceriainty by means of a few observations upon its 

 wells and springs, made within the compass of a single year, that it 

 seems a very useless waste of time and attention to watch, and record 

 with reference to this object, the indications of a thermometer ex- 

 posed to the air. 



The columns for the direction of the winds, for the aspect of the 

 heavens and the amount of rain, that appear in the best meteorologi- 

 cal registers are not without their value, but their usefulness would 

 be very much increased if measures were instituted for effecting an 

 immediate comparison of them with the observations made in other 

 and distant parts of the country. When observations with the hy- 

 grometer shall have been sufficiently multiplied to warrant us in 

 drawing conclusions from them, it will probably appear that the quan- 

 tity of vapor is less on the western than on the eastern side of the 

 Atlantic — that though the amount of precipitation is greater the air 

 is considerably drier. 



It is however by a minute and thorough study of individual phe- 

 nomena, and tracing the progress of the changes that occur from 

 time to time over a. wide extent of country, that the science of me- 

 teorology is to be perfected. The truth of the proposition stated by 

 Leibnitz — " JVeque aliud est natiira quam ars quaedam magna,'^ is 

 to be borne constantly in mind. Every great storm is a succession 

 of chemical changes, immense alike in their number, magnitude, and 

 the space through which they occur, and as it is by a close attention 

 to all the circumstances of single experiments, and not by a loose 

 and indefinite approximation of a number, that chemistry has made 

 such astonishing advances, so meteorology, if studied at all with suc- 

 cess, must be pursued in the same way, the labors of many different 

 persons being combined in the collection of data, where the observa- 

 tions of a single individual would be inadequate to the attainment of 

 the object in view. 



It is not known whether the plan sketched in the following letter 

 of Professor Brandes, dated Breslavv, December 1, 1816, copied 

 from Gilbert's Annals into the Bibliotheque Universelle, and transla- 

 ted from thence for the Journal of Science, has been carried into 

 execution, but it will be apparent that it might be applied with ad- 

 vantage to our own country. 



" Some investigations which I had proposed to attach to this letter, 

 have not afforded the results I expected from them. I had collected 



