2/2 SHOOTING STARS. 



radiant points are on the equator) the activity may extend itself to 

 more than a month. I find, however, that not only is there a 

 second outbreak of meteors, but that many of these radiant points 

 endure for five or six weeks at least, and that they are stationary 

 amongst the stars, which is also another and important objection 

 to our received theories. To account for such anomalies is 

 extremely difficult, and it has been sought to explain the re- 

 appearance of these meteor showers, on the assumption that they 

 are quite different systems coming from the same directions, and 

 that in fact the agreements are purely accidental. Such an explana- 

 tion will not hold however, because the positions are exactly the same. 

 There are however so vast a number of meteor streams differing in 

 dates and positions, that it is only reasonable to expect that many of 

 them should overlap or overlie, and that thus, by the incorporation of 

 several distinct systems, a false duration is brought about. But 

 allowing for all such objections, and giving errors of observation a 

 wide margin, (for in this special department it is impossible to be 

 absolutely exact in dealing with such transient and intermittent 

 bodies,) I cannot see my way clear to account for the long duration 

 of many showers. The number of instances is so large and so 

 sharply defined that it is impossible to disconnect them. In this 

 relation it is of extreme importance to note the special 

 peculiarities of the meteors belonging to each shower. Some 

 systems give slow members, others very swift, and others again 

 will be accompanied by streaks or trains. Another family will 

 consist wholly of short meteors, while yet another will exhibit long 

 sweeping courses. Now when these individual characteristics are 

 carefully noted, it is possible to sort out the meteors to their several 

 radiant ooints, and to group them into separate systems according 

 to their distinctive visible appearances. 



If these long continued displays to which I have been referring 

 each belong to one and the same stream, it is reasonable to expect 

 that the meteors of each should possess features in common, and 

 this I have really found to be the case in a few instances, and I 



