ON SHOOTING STARS. 21 



The velocities have been considered as uniform. But it is evident thai if si. me shooting 

 stars have larger and some smaller velocities, the hourly distribution ought to be approxi- 

 mately that correspond in-- the mean velocity. It has been assumed, moreover, that the 

 number seen is proportional to that of those coming from all parts of the visible heavens. 

 But if the number seen is proportional to the number that strike the upper surface of the 

 atmosphere within the circumference of a given circle of moderate dimensions, the centre of 

 the circle being in the zenith of the observer, then must the number coming from each 

 element of the visible heavens be multiplied by the cosine of the zenith distance of that 

 element. By this supposition the hourly nuuTbers would be increased in the morning and 

 diminished in the evening from those given in table IV. Neither of these .suppositions is 

 probably strictly correct. 



It should be borne in mind, on the other hand, that the earth's attraction tends to make 

 the numbers more uniform through the night. There should also be less difference in the 

 hourly numbers during the last half of the year than in the first half, since the point to which 

 the earth is moving is then north of the equator. 



If now we compare the numbers in table IV with Mr. Herrick's estimate, and with Mr. 

 Coulvier-Gravier's hourly numbers given above, a mean velocity of the shooting stars is 

 indicated as large as, or larger than, that of comets in parabolic orbits. The nature of the 

 data will nfB, however, allow this conclusion to be strongly urged. Yet that the mean 

 velocity is greater than that of the earth in its orbit seems almost certain. 



NUMBER OF METEOEOIDS EST THE SPACE WHICH THE EARTH IS TRAVERSING. 



We have found that, of the space through which the earth is travelling, a volume equal 



to that of the earth (atmosphere included) contains a mean number of meteoroids expressed 



by the equation, 



RV 



M= 116.2 



V 2 +v*+0v 6 



In this equation Vis the mean relative velocity of the meteoroids. If the absolute 

 velocities were all uniform, and the points from which they come uniformly distributed over 

 the heavens, we should have evidently, 



p* i _ ,_■" 



V = A I (i! ! +ii B -2s»' cos w) sin wdw — v + k — . 



Jo . ■» 



For v' we may use, as an approximation, the mean absolute velocity. If v' = v, then V = fi>. 

 If v' = Vi/2, then V=|». As the mean velocity of the meteoroids seems to be greater than 

 that of the earth, and more nearly equal to that of bodies moving in parabolic orbits, the 

 latter value will be used for V. Computing M on this supposition, neglecting the term v a \ 

 we have its value more than 14,000. If we deduct for the volume occupied by the atmo- 

 sphere, we have more than 13,000 bodies in each volume of the size of the earth, each body 

 such as would furnish a shooting star visible under favorable circumstances to the naked eye. 

 40 (311) 



