264 PROCEEDINGS OP THE AMERICAN ACADEMY 



metres, the focal lengths one metre, and the height of the cloud one 

 kilometre ; then the difference between the distances of the two lines 

 will be a decimetre. K this distance is measured with an error no 

 greater than a millimetre, the height will be given wdthin ten metres, 

 or within one per cent. The velocity per minute is then readily 

 deduced from the lines perpendicular to the direction of the wind, and 

 the velocity of the latter may thus be determined within one or two 

 per cent, a degree of accuracy at least equal to that of the best deter- 

 minations of tlie velocity of the wind at the earth's surface and much 

 greater than the degree of uniformity of any ordinary wind. Each 

 cloud will furnish a measurement at a different height, and a compari- 

 son with observations at the surface of the earth will readily give the 

 relative velocities at these various altitudes. 



Various other applications of this principle will suggest themselves. 

 For instance, if the paper is replaced by a sensitive photographic 

 2:)late, and the cameras directed towards a distant tliunder cloud at 

 night, an image of each flash may be taken. A great many flashes 

 may be recorded on each plate, and the corresponding images recog- 

 nized by their forms. The distances and true dimensions may then be 

 determined with considerable accuracy. If observations are made at 

 the same time, of the interval between the flash and the thunder, the 

 velocity of the sound may be measured, and it may be proved whether, 

 as has been claimed, the velocity of such an intense sound is far greater 

 than that of any ordinary noise. 



