Beam of November 17, 1882. 325 



observations. And, first, (column 4), time of its appearance. 

 Owing no doubt to the sudden and unexpected advent of the 

 object, the times recorded are evidently not very certain, and, 

 with the exception perhaps of the Greenwich and some of the 

 foreign observations, are doubtless neither chronometer nor 

 clock time. Many do not profess to be exact; and we are not 

 generally told whether they are Greenwich or local, though 

 probably they are the former. I think, however, they are 

 sufficiently near to enable us to identify observations 1 to 21 

 and 24 to 28 as all applying to the same phenomenon. For any 

 other purpose the recorded times do not seem of much prac- 

 tical use. Secondly, time occupied in transit (7th column). 

 This is also involved in some obscurity. One observer (obs. 12) 

 gives so short a time as " about 6 seconds/'' while another 

 (obs. 10) extends the time to a length of "less than 4 

 minutes ;" and between these rates we get estimates mostly of 

 1 and 2 minutes, accompanied by such qualifying expressions 

 as "less," "about," &c. A mean of eighteen observations 

 is found to be 1 minute 15 seconds. As to the longer periods, 

 we know by the experience of counting beats by clock or 

 chronometer that the estimate of time by observers not so 

 accustomed is usually an overrate ; and I think we may con- 

 sider this to have been the case in these instances. Thirdly, 

 as to apparent length and breadth of the object (columns 8 

 and 9). These of course, as with the time occupied in transit, 

 would vary with the position of the observer ; and at places 

 situated more southward the object would have been seen of 

 greater length and breadth than when seen from the north, 

 while east and west positions would get foreshortened effects. 

 Perspective would come into play with the position of the beam 

 in the sky, most observers agreeing that it lengthened out as it 

 neared the zenith, though one (Mr. Batson, obs. 4) thought it 

 contracted as it passed across the moon. The observations, 

 however, are really not conformable; and though three 30° 

 make their appearance, the observers (obs. 5, 9, 14) differ in 

 position and in their time of transit estimates. The estimated 

 lengths, as might be expected, differ more than the breadths. 

 Perhaps the two most inconsistent records are the Peruwelz 

 (obs. 26) and the Bruges (obs. 27) observations with the same 

 breadth but so great a difference in length. A mean of eleven 

 observations for length (excluding the Rye and Peruwelz 

 observations) and of nine observations for breadth gives 27° x 

 3 J° as a general idea of the beam's apparent size. 



Column 10 brings us to the direction of flight. In this 

 case we have more actually accordant observations, inasmuch 

 as all agree in general terms upon a flight from an east to a 



