ON THE ELECTRIC COLUMN. gfil 



successively augmented upon its original base, I suspect 

 that it is too much crowded, and that thus its parts may- 

 have on each other an influence prejudicial to the effects, 

 which I have remarked in other cases. 



I have made many other remarks on this instrument, but 

 my present purpose is more to engage other experimental 

 philosophers in this pursuit, than to forward it myself ; for 

 with re'ipect to these observations, I consider them as newly 



born. It will first require some time for the understanding ^^^'"^'*^^*'' 



„ , . . . ons of the in- 



what may be called the language oi the mstrument ; i. e. its strument to be 



meaning as to the indication of the electric state of the^^"^^^^* 

 ambient air, by its influence on the motions of the pen" 

 dulum. This study has been opened too late for me, 

 though I was engaged in it by considerations resulting from 

 long meteorological observations, which, as they are of the 

 greatest importance to natural philosophy, must be the 

 incitement to this pursuit. Wishing therefore, that such 

 observations may become a more general object of atten- 

 tion among natural philosophers, I have here endeavoured 

 to show, by an abstract view of their present results, what 

 knowledge, in following them, may be still obtained con- 

 cerning the atmospheric operations. It is true, that obser- 

 vations of this kind require the neighbourhood of moun- i^,;„|j p^^^g ^f 

 tains (unless those who ascend in balloons should carry '^'^^ ^t*"*** 

 proper meteorological instruments, and apply themselves to ^^ examined 

 these observations). But in general no real knowledge of 

 the nature of the atmosphere can be obtained without, in 

 some manner, ascending in it ; and it is no less certain, 

 that without this knowledge, no chemical theory can possess 

 any certainty. 



Systems are useful for promoting science, provided they Systems use- 

 be founded on all the knowledge already acquired respecting 

 their object ; but even then, as long as they contain hypo- 

 theses, they must be only considered as leading to new 

 researches on determined points. With this view, I shall 

 here conclude by an abstract of a meteorological system 

 which I have fully explained in my former works, and espe- 

 cially in that under the title of Introduction a la Physique 

 terrestre par les Fluides expan&ihles, 



I. During 



