ON SnOOTING STARS- 6 



The observations made in this country in August and November of last ) r ear, and the more 

 reliable series of European observations, lead me to doubt very much the smaller altitudes. 

 Most of them are in fact computed from very short b*se lines. While there maybe some 

 flights which are quite low, I feel safe in disregarding all the cases where the height of tho 

 middle point is said to be less than thirty kilometres. On the other hand, there must be a 

 definite upper limit to the appearance of these trains, and the extension of the curve given 

 above to the right of 180 at least, may be referred to errors of observation. The numbers 

 corresponding I shall therefore disregard. In the following computations the rejection of 

 these very high altitudes produces an effect the opposite of that which results from rejecting 

 the very low ones, and the two effects tend to balance each other. 



I shall assume that these observed paths are fair examples as to altitude of all visible 

 paths, and hence that the frequency of the middle points at different altitudes above the 

 earth's surface is proportional to, and may be expressed by, these numbers: 



From 30 kilometres to 60 by 114 



GO " 90 " 243 



90 " 120 " 277 



" 120 " 150 " 10G 



" 150 " 180 " 57 



Representing these numbers by p, and the average altitude by h , we have, approximately, 



where in the finite summation indicated by 2 the successive values of p are to be taken, and h is 

 to be successively i (30 + GO), ± (GO + 90), &c, or 45, 75, 105, 135, and 1G5 kilometres. The 

 value of h , that is, the average altitude of the middle points of the meteor-paths above the 

 earth's surface, is thus found to be 95.5 kilometres, or 59.4 English miles. 



■DISTRIBUTION OF THE METEOR-PATHS OVER THE APPARENT HEAVENS IN AZIMUTH. 

 If the middle points of the apparent paths of all the shooting stars that can be seen at one 



place during a long time were marked on the visible heavens, we jr z J- 



might reasonably suppose these points equably distributed in azimuth. 



For let O, in figure 2, be the place of an observer, Z his zenith, 



and let the paths, or their middle points, be distributed through a dS" 



stratum above him. If then OA and OA l make equal angles with !° 



OZ, these lines will pass similarly through the stratum, and about -^ - rf^== = — 



as many paths should be seen along one line as along the other. }}/ 



This argument implies an equable distribution of the paths horizontally along the 



stratum, which is not strictly true. For the number of meteors seen at a place increases 



through the night, and hence there should be east of us more paths than west of us at any time 



of the night. The difference, however, is quite small. A uniform direction of tiie paths may 



also have a little influence, especially by changing the distance at which paths can be seen. 



(293) 



