252 



NATURE 



[July 12, 1888 



tions. In fact three of the four changes just named 

 make the evidence for my conclusions weaker instead of 

 stronger. 



In the treatment of the observations several quantities 

 have been neglected as not large enough to be comparable 

 with the probable errors of the observations themselves. 

 Thus the effect of the earth's attraction in changing the 

 direction of motion, or what has been called the zenithal 

 attraction of the quit, has been allowed for only in a 

 general way. So the earth's quit and goal are treated as 

 being exactly 90° from the sun ; or, in other words, the 

 earth's orbit has been treated as a circle. In like manner 

 the motion of the place of fall due to the earth's rotation 

 on its axis has not been taken account of. 



Having located upon the chart the meteor's relative 

 quit we have next to construct its absolute quit. This 

 evidently lies on the great circle joining the relative quit 

 to Q (Fig. 1), which, when the sun is at S is represented 

 on the chart by a straight line through Q, together with 

 its corresponding line through G. When the absolute 



velocity of the meteroid in its motion about the sun is 

 given, the place on this circle of the absolute quit can be 

 determined by combining by the parallelogram of velocities 

 the motions of the earth and of the meteoroid. The 

 following table is an abstract of a larger one used in this 

 reduction, and is constructed for the limiting velocities 

 1 '414 and 1 '244 : — 



Table showing the Distances from the Earth's Quit to the Absolute 

 Quit of a Meteoroid for Different Distances from the Earth's 

 Quit to the Relative Quit of the Meteoroid. 



Distance from Q to relative quit. Distance from Q to absolute quit. 



v = 1 '414. z>-= 1-244 



3o° 9° - 3 6° -3 



60 22 - i 15*8 



90 45-0 36-5 



120 82-1 75-8 



150 129-3 126-3 



180 1800 1800 



In the following constructions the maximum velocity of 

 the meteoroid has been used. When the meteoroid's 



Fig. 1. — Showing the distribution of 116 meteorite quits relatively to the sun's place and to the earth's quit. 



relative quit is known as a point the absolute quit is at 

 once constructed. If, however, we have an area within 

 which the relative quit is probably located we may mark 

 off with equal facility points on the boundaries of the 

 area within which the absolute quit is probably located. 

 If the former area is a circle the latter will be an oval. 

 The centre of the circle does not correspond exactly to 

 the centre of the oval, but by applying a correction to the 

 table the centre of the oval absolute quit area can be 

 directly constructed from the centre of the circular 

 relative-quit area. 



In Fig. 1 I have given in a single diagram constructed 

 on a stereographic projection, the results for 1 16 stone-falls. 

 The best determinations which the accounts admit of for 

 the meteor's direction were first made out. Then the 

 centre of the probable quit area in each case was assumed 

 to be the actual quit. When only the quarter of the 

 heavens from which the stones came is stated the centre 

 of probable area was taken 20 above the horizon. Inter- 

 preted thus, the stars in Fig. 1 represent the places of the 



1 16 absolute quits relatively to the place of the sun, S, and 

 to that of the earth's quit and goal, Q and G. 



Let us denote any one of these quits (or stars), by the 

 letter q. The elements of the orbit in which the 

 corresponding stone was formerly moving can be easily 

 obtained from the projection. The earth's longitude on 

 the day of fall is the longitude of the node. The angle 

 ^SQ is the inclination of the orbit to the ecliptic, and its 

 amount is at once read off on the projection. The orbit 

 has been assumed to have been a parabola. Hence, twice 

 the complement of ^S was the angular distance of the 

 stone from its perihelion. If ^S > 90°, the perihelion 

 had not been reached ; if ^S < 90, the perihelion, had 

 been passed. The perihelion distance was sin 2 QS. If, 

 however, it be assumed that the orbit was a long ellipse 

 of given major axis, the place of the absolute quit, q, 

 moves somewhat nearer to Q along the line^Q, the angle 

 in the plane of the orbit from perihelion was a little more 

 than twice the complement of ^S, and the perihelion 

 distance somewhat less than sin'^S. But all these 



