4 . ON SHOOTING STARS. 



The above conclusion accords with observation. Mr. Hekrick was accustomed to watch 

 in company with others on the anniversaries of the August and November showers, and some- 

 times on other nights. He classified the paths carefully according to the quarter of the 

 heavens in which they originated, sometimes dividing the heavens into the N., E., S., and W. 

 quadrants, at other times into N.E., S.E., S.W., and N.W. quadrants. His results, as well as 

 those of others observing in the same manner, are published in the volumes of the American 

 Journal of Science since 1837. Rejecting all the observations when the heavens were not 

 divided into four parts, and collecting the remainder, we have of 6,598 observed paths — 



800 in N. 733 in N.E. 



965 in E. 852 in S.E. 



847 in S. 833 in S.W. 



889 in W. 679 in N.W. 



These numbers imply a small predominance in the southeast. Since the average zenith 

 distance of the middle points of paths is 48°.3, the above numbers give as a centre of gravity, 

 or centre of distribution, a point about 2° from the zenith, in the direction S. 28° E. Mr. 

 Herrick classified them according to the place from which they proceeded. This makes the 

 centre of distribution a little nearer the zenith, and perhaps more to the east, than it would 

 have been if fie had classified them according to their middle points. 



Mr. Coulvier-Gravier has given in his Reclierches sur les Etoiles Filantes some deductions 

 from 2,309 paths observed by him. A table given on page 184 implies that a point about 5° 

 from the zenith, and very nearly northeast, is the centre of distribution. This is farther from 

 the zenith than the point before given. The observations are of one person, and are of course 

 affected by his habits of watching. Combining both results we have the centre a little south 

 of east of the zenith, and 1°.6 distant. We may safely regard these deviations from the 

 zenith as errors of observation, and consider the relative frequency of occurrence of meteor- 

 paths in different parts of the visible heavens as a function of zenith distance only. 



DISTRIBUTION OF METEOR-PATHS AS TO APPARENT ALTITUDES. 



We have seen that the relative frequency of meteor-paths in different parts of the visible 

 heavens is a function of zenith distance only. The nature of that function cannot be deter- 

 mined a priori. But that there is a very rapid diminution of brilliancy as we approach the 

 horizon is thus shown. 



Let m and m 1 in the foregoing figure be the places of two meteor-paths of equal intrinsic 

 brightness, and equal altitude, one seen in the zenith from 0, and the other at a distance 6 

 from the zenith. The one at in 1 will appear less bright at than that at m, both because it 

 is farther from 0, and because its light has to traverse obliquely the stratum of air and mists 

 which is near O and much below the region of shooting stars.*" Since Om 1 cos 6 = O/;;, the 

 diminution of light from distance is expressed by the factor cos 2 0. 

 (201) 



