I 



SHOOTING STARS. 2^ \ 



equally distributed over the 12 months. August and October 

 appear to give an abundancy of shooting stars, while the spring 

 months, especially March, give them in scanty numbers. These 

 deductions must hovi^ever, be regarded as approximations that 

 require more proof before acceptance. They are merely the best 

 construction that I can put upon my necessarily incomplete 

 observations, and no doubt will require revision when our 

 knowledge of these phenomena is more advanced, and they are 

 watched more generally and systematically than at present. 



The number of meteors actually projected by me on star charts, 

 including those observed and those selected from published 

 catalogues, reaches over 10,000 • but, in addition to this, the paths 

 of fully 20,00c others were examined, to see if their directions were 

 such as to conform to certain new showers that it was desirable to 

 support. Such a varied investigation, and one so extensive, has 

 never, I believe, been undertaken before, (in fact the materials did 

 not exist,) and an analysis of the results has opened up some 

 interesting questions. The most prominent of these is that many 

 of the minor long-continued showers apparently occur twice over, 

 that is to say, there is a second outbreak of meteors from exactly 

 the same points after a period of apparent cessation, and 

 that at intervals of three months the two showers apparently attain 

 their maxima. This is a rather startling announcement to those 

 who understand the theory of these shooting stars, and is equivalent 

 to saying that a person may, by looking towards the eastern sky in 

 the evenings of November, see the very same showers come into 

 play that were visible in the mornings of August from the same 

 eastern constellations. Now such a thing is wholly opposed to our 

 present ideas, as to the astronomical nature of these phenomena. 

 Their orbits, as we have been given to understand, being the same 

 as comets, will only, in the great majority of cases, intersect the 

 earth's path once annually, so that the same shower can in nowise 

 break out again until a similar date each year, and the duration 

 cannot exceed a few nights except in exceptional cases (^when the 



