April 7, 1898] 



NATURE 



541 



the ordinary shooting-stars, and their heights at extinc- 

 tion appear to be about 30 miles. For the present 

 purpose it is intended to refer to the elevation of these 

 objects at the beginning of their visible flights, for this 

 elevation is so considerable in some cases, that, if atmo- 

 spheric friction induces their combustion, the air extends 

 to a much greater distance from the earth than is 

 ordinarily supposed. 



It is not at all a rarity to find meteors which, at the 

 instant of their first appearance, were more than 100 miles 

 in height. I have looked through various lists of the 

 computed real paths of fireballs and shooting-stars, and 

 find that, out of 577 cases, 1 16 exhibited a beginning 

 height of 100 miles or more, the average being 130 miles. 

 In fact, one meteor out of five displayed incandescence 

 when 100 miles or more from the earth's surface. The 

 materials from which I obtained these results were by 

 Dr. E. Heis, Prof. A. S. Herschel, Prof G. von Niessl, 

 and myself The most extreme heights ^ were : — 



Date of meteor. 



1868 September 5 

 1849 August II ... 



1861 July 16 



1862 February 2 

 1864 August ID ... 

 1883 June 3 



1861 August 10... 

 1864 July 28 ... 

 1870 September 27 

 1877 March 21 ... 



Height 



at beginning. 



Miles 



.. . 483 ... 



... 216 ... 



... 195 ••• 



... 190 ... 



... 188 ... 



... 188 ... 



... 184 ... 



... 184 ... 



... 184 ... 



... 184 ... 



Authority. 



G. von Niessl. 



E. Heis. 



A. S. Herschel. 



E. Heis. 



G. von Niessl. 



E. Heis. 



G. von Niessl. 



The first of these is probably erroneous, for the observ- 

 ations, though numerous, were not accordant, and with 

 such data it is possible for different computers to work 

 out anomalous results. Thus, in the instance of the very 

 long-pathed fireball seen in France and Germany in 1868, 

 three paths have been computed, and they differ widely 

 in their character. These differences are induced by the 

 erroneous observations, and the difficulty of putting a 

 consistent interpretation upon them. The radiant point, 

 as adopted by the various computers, is dissimilar ; and 

 this in itself must occasion a great discordance in the 

 heights, for one observer putting the radiant 5° above the 

 horizon will obviously obtain a lower elevation for the 

 beginning point than another who places it 15° above the 

 horizon — the angle of the meteor's descent being much 

 less. In regard to the fireball of September 5, 1868, the 

 following results were obtained : — 



Thus, while von Niessl made it descend from 483 to 

 115 miles, M. Tissot concluded that it really ascended 

 from 69 to 191 miles ! Prof Herschel's results appear to 

 be the best that can be derived from the materials avail- 

 able, for he obtains normal heights and a slight ascent 

 of the meteor just before extinction. Its enormous length 

 of path is quite beyond dispute. 



In every instance where the observations are very in- 

 consistent, it is clear that the results of investigations of 

 this kind must depend largely upon the interpretation 

 put upon them. And for strictly scientific purposes the 

 real paths derived from such materials are of little use, 



_ 1 Other instances of abnormal height might be quoted from the deduc- 

 tions of other authorities, but they are open to serious question. Thus, for 

 the fireball of March 19, 1718, the height at first appearance has been given 

 at 297J miles ; but Prof Herschel finds, from a careful rediscussion of the 

 observations, that the meteor began at an elevation of only 80 miles. 



* This is the lowest elevation of the meteor as found by M. Tissot, and 

 quoted in British Association Report for 1869, p. 272. 



for any critical deductions or trustworthy comparisons 

 cannot be made from them. The instance above alluded 

 to furnishes, however, a very exceptional case ; but it has 

 been selected in proof of the great uncertainty attaching 

 to deductions based upon conflicting observations. 



It appears that about 20 per cent, of meteors are at 

 least 100 miles high at the instant of their first visible 

 apparition. This conclusion rests upon a considerable 

 number of results, including a large proportion of fire- 

 balls, and may be trusted within small limits of error. 

 From the materials I have examined, I believe the actual 

 height at first appearance of a meteor is very rarely as 

 much as 150 miles, and that it seldom reaches beyond 

 130 miles. . 



It is singular that in 1897 I found unusual elevations 

 for several meteors, in fact 9 out of 26 {i.e. more than 

 one-third), whose real paths I computed, indicated a 

 beginning-height of over 100 miles. These were : — 



Date, 1897. 



h. m. 



Aug. 2, II 54 



2, II 24 



8, 9 15 



9, 13 27 

 9, 13 52 

 9, 14 18 



Nov. 13, 15 28 



13, 15 52 



Dec. 12, 8 6 



Mag. 

 2 



5-4 



> ? 

 3—1 



3 



3x ? 



I 



I 



> ? 



Height 

 at be- 

 ginning. 

 Miles 

 .. 112 . 



• 139 

 - 133 

 ... 140 



... 131 



... 137 



... 125 



... 103 



... 112 



Height 



at 



ending. 



Miles 



,. 90 . 



.. 124 . 



.. 115 . 



.. 77 . 

 .. 89 . 

 - 75- 



.. ^^ . 



■■ 59 ■ 

 .. 19 . 



Length 



of 



path. 



Miles 



. 40 ., 



. 28 . 



. 63 . 



. 81 . 



. 56 . 



■ 75 • 



■ 75 • 

 . 60 . 

 . 151 . 



Radiant. 



40 + 55 

 73 +66 

 52+47 

 46 -I- 56 

 58 + 60 

 44+45 

 136+ 9 



152 -F 22 



80 + 23 



It is possible that in several of these cases mistakes of 

 identification may have occurred. It must sometimes 

 happen, and especially during the occurrence of a rich 

 shower, that two meteors are recorded at the same time 

 at different places, which show parallax in the right 

 direction, though they are entirely separate objects. 

 Accidental coincidences of this kind would, however, not 

 very often occur, and they would usually be detected by 

 some features of mutual discordance. 



There is another point in connection with the first 

 appearance of meteors which merits attention — this is, 

 that observers seldom secure an accurate view of it. The 

 end point is more precisely determined as the eye steadily 

 follows the object until its extinction. But it is rarely 

 the case that even an habitual observer of meteoric phe- 

 nomena happens to be looking directly to that point of 

 the heavens where a meteor appears. He generally 

 catches it after it has already traversed a section of its 

 flight, and often estimates the extent of its backward 

 trajectory, sometimes adding 5° or 10° to the observed 

 starting-point. Now, a slight error in carrying the 

 visible line of flight too far back may put 30 or 50 miles 

 on the beginning-height of a meteor, especially if it is 

 anywhere near its radiant. It would, therefore, be safer 

 for observers to record the path actually witnessed, with- 

 out assuming the extent of the portion which escaped 

 them. 



But apart from all the uncertainties (which have their 

 outcome in the rough character of the observations) 

 attaching to the subject, it is impossible to put aside the 

 evidence that meteors are sometimes 130 miles and, in 

 extremely rare instances, 150 miles high when they are 

 first visible. There are grave doubts that any meteor has 

 ever been visible at a height of 200 miles. And it is 

 probable that many, if not all, of the instances where 

 heights of about 170, 180 or 190 miles have been found, 

 were due to the commencing points of the flights having 

 been carried too far back by the observers, or that mis- 

 takes in the directions have led the computer to adopt 

 erroneous radiants and deduce initial heights consider- 

 ably in excess of the correct ones. 



If photography could step in here, and dispel all the 

 doubts arising from our hurried and often questionable 



NO. 1484, VOL. 57] 



