1910] on Bailey's Comet. 761 



10,000 comets under the same circumstances to provide us with one a 

 jear. Now, we scarcely know enough about comets in general to know 

 what the average period is, but it is probably something in the 

 thousands of years, and there must therefore be thousands of comets, 

 which spend most of their time at a distance from the sun, hanging 

 between successive journeys to him, while there may be millions — 

 our knowledge is too imperfect to guide us. We can, however, assign 

 a superior limit in some such way as this : if a comet travelled more 

 than half way to the nearest star (supposed equal to the sun) it would 

 be hable to desert the sun for the star. We may therefore take some- 

 thing like 100,000 times the earth's distance from the sun as a limit 

 for the average major axis of a comet's orbit, giving a period of about 

 10 million years. Since we see about M comets of long period per 

 year and we may miss several, there may be (say) 50 million comets, 

 but there are not hkely to be more, assuming them permanent members 

 of the solar system.* In forming a mental picture of the universe 

 we must not forget to include this possible envelope of comets sur- 

 rounding each star. 



Messrs. Cowell and Crommehn have not limited their work to 

 the present return of the comet, but have carried its history back 

 through the ages to 240 B.C. Nearly a century after Halley's death 

 a tine piece of work in continuation of his great discovery was 

 accomplished by Mr. J. R. Hind, who, by examining old records 

 and especially the Chinese Annals, was al)le to indicate with fair 

 probability the following previous appearances of Halley's comet : — 



Peobable Eaely Returns of Halley's Comet (Hind) 



But a more complete discussion was needed : several of the 

 identitications were uncertain, and one or two of them turn out to 

 be wrong. Messrs. Cowell and Crommelin, with the assistance of 

 three volunteers, Dr. Smart, Mr. F. R. Cripps, and Mr. Thomas 

 Wright, have proceeded backwards step l)y step, making each return 

 sure before proceeding to the next. B.C. S7 and (probably) B.C. 240 

 have been added to the list of returns. Hind was wrong in 608 

 (a year and a half too late) and in 912 (four months too early) and 

 in 122;; (ten months too late). The comet of 1222 was a bright 

 one, seen ])oth in Europe and China, but its identity with Halley's 

 was not suspected until this careful investigation was made. The 

 date 1066 wiU be noticed as that of the Norman Conquest of 



* Dr. Lardner in a diSerent manner estimated 7 millions. 



