380 M. Sccchi on Shooting -Stars. 



the construction gave a height nearly identical for both stations, 

 amounting to 105 kilometres. 



Here, then, is a new and unexpected difficulty in obtaining 

 exact results. Its solution, nevertheless, is not difficult. It will 

 suffice in fact to apply in such cases the principles of descrip- 

 tive geometry, and to trace the direction of the real trajectory 

 according to any two portions of the apparent trajectories seen 

 from the two different stations. Prof. F. Armellini kindly 

 undertook this determination in many of the cases which proved 

 most difficult and rebellious to the simple geometrical construc- 

 tion. His results were very satisfactory ; it was thus proved 

 that these cases followed the rule applicable to the others, but 

 it was necessary to admit that the star had become visible at 

 different times from the two stations. Another consequence of 

 these more accurate constructions is the approximate correctness 

 of the mean of the two heights deduced from the simple con- 

 struction — at all events when the latter do not differ too greatly. 



This premised, I proceed to the results obtained for the 

 heights of the meteors. In round numbers the mean of all 

 these heights was from 101 to 100 kilometres. There were, 

 in fact, 



6 stars whose heights were between 40 and 60 kilometres. 



7 „ „ „ 60 „ 80 „ 

 10 „ „ „ 80 „ 100 „ 

 17 „ „ „ 100 „ 120 „ 



3 „ „ „ 120 „ 140 



5 „ „ „ 140 „ 160 „ 



2 „ „ „ 160 „ 180 



1 „ „ „ 180 „ 200 



2 „ „ „ 200 „ 220 



3 „ „ „ 200 and upwards. 



Thus the heights of 27 out of the whole 56 stars (about 50 

 per cent.) fell between 80 and 120 kilometres. As in many 

 cases the trajectory was perpendicular to the vertical plane pas- 

 sing through the two stations, the height of the trajectory was 

 constructed directly by the principle of parallaxes, and amounted 

 to 93 kilometres. The mean of the heights of the points of 

 extinction was 75 kilometres. The greatest of the observed 

 heights amounted to between 240 and 260 kilometres ; but it 

 must not be forgotten that these values may have been influ- 

 enced by the errors of observation above discussed. Neverthe- 

 less in these cases the heights must have exceeded 200 kilo- 

 metres. 



