110 CELESTIAL PHENOMENA. 



the river near Narni, projected nearly four feet above the sur- 

 face of the water. It must be remarked that these meteoric 

 stones, whether ancient or modern, cannot be regarded as 

 more than principal fragments of the mass which exploded 

 in the fire-ball, or descended from the dark cloud. 



When we consider the mathematically-proved enormous 

 velocity with which meteoric stones arrive at the earth from 

 the extreme limits of the atmosphere, and with which balls 

 of fire move in a more lengthened course through its denser 

 strata, it appears to me highly improbable that these metal- 

 liferous masses of stone, with their imbedded and perfectly- 

 formed crystals of olivine, labradorite, and pyroxene, should 

 have condensed from a gaseous state into a solid nucleus in 

 so short an interval of time. Even when the falling pieces 

 differ from each other in chemical composition, they almost 

 always show the peculiar characters of a fragment, having 

 often a prismatic, or truncated pyramidal form, with slightly 

 curved faces and rounded angles. But whence this frag- 

 mentary character (first recognised by Schreibers) in a 

 rotating planetary body ? Here, as in the sphere of organic 

 life, there is obscurity in all that belongs to the history of 

 development. Meteoric masses kindle and become luminous 

 at elevations which must be supposed to be almost 

 entirely deprived of air. Biot's recent investigations 

 on the important phenomena of twilight ( 63 ) even reduce 

 considerably the height of the line which has been 

 usually, but somewhat hazardously, termed the " limit" of 

 the atmosphere; but luminous processes may take place 

 without the presence of oxygen, and Poisson has imagined 

 | that the ignition of aerolites occurs far beyond the range of 

 \ our atmosphere. In treating of meteoric stones, as well as 



