438 Suppcfed Height [Book V. 



indeed, that the vaft magnitude of fuch bodies might 

 compenfate for the thinnefs of the medium in which 

 they moved j whether or not this was the cafe, cannot 

 indeed be afcertained, as we have fo few data to go 

 upon; but the greateft difficulty is to account for the 

 brightnefs of the light. Appearances of this kind are, 

 indeed, with great probability, attributed to electricity, 

 but the difficulty is not thus removed; though the 

 electrical fire pervades with great eafe the vacuum of 

 a common air-pump, yet it does not in that cafe ap- 

 pear in bright well defined fparks as in the open air, 

 but rather in long ftreams refembling rhe aurora bo- 

 realis. From forne late experiments, Mr. Morgan 

 concludes that the electrical fluid cannot penetrate a 

 perfect vacuum. If this mould be the cafe, it fhews 

 that the regions we fpeak of are not fuch a perfect 

 vacuum as can be artificially made ; but whether they 

 are or not, the extreme brightnefs of the light fhews 

 that a fluid was prefent in thofe regions, capable of 

 confining and condenfing the electric matter as much 

 as the air does at the furface of the ground -, for the 

 brightnefs of thefe meteors, considering their diflance, 

 cannot be fuppofed inferior to that of the brighteft 

 flames of lightning. 



' It appears, therefore, that the abfolute height of 

 the atmofphere is not yet determined. The begin- 

 ning and ending of twilight, indeed, mew, that the 

 height at which the atrrofphere begins to refract the 

 fun's light is about forty- four or forty-five Englifh 

 miles. But this may, not improbably, be only the 

 height to which the aqueous vapours are carried ; for 

 it cannot be thought any unreafonable fuppofition, that 

 . light is refracted only by means of the aqueous va- 

 pour contained in the atmoiphere: and where this 

 ceafes, it is dill capable of fupporting the electric fire 



at 



