576 COSMOS. 



which are always represented, others which appear only 

 sporadically and changeably." 



Whether, moreover, the different points of divergence alter 

 with the years which, if closed rings are assumed, would indi- 

 cate an alteration in the situation of the ring in which the 

 meteors move cannot at present be determined with cer- 

 tainty from the observations. A beautiful series of such 

 observations by Houzeau (during the years 1839 to 1842) 

 appears to offer evidence against a progressive alteration. 1 * 

 Edward Heis lf has very correctly remarked that in Grecian 

 and Roman antiquity, attention had already been directed to 

 a certain temporary uniformity in the direction of shooting 

 stars darting across the sky. That direction was then con- 

 sidered as the result of a wind already blowing in the higher 

 regions of the atmosphere, and predicted to the sailors an 

 approaching current of air descending thence into the lower 

 regions. 



If the periodic streams of shooting stars are distinguished 

 from the sporadic by the frequent parallelism of their paths, 

 proceeding from one or more points of divergence, a second 

 criterion of them is the numerical the number of individual 

 meteors referred to a definite measure of time. We come 

 here to the much-disputed question of the distinction of an 

 extraordinary from an ordinary fall of shooting stars. Two 



118 Saigey, p. 151 ; and upon Erman's determination of the 

 points of convergence diametrically opposed to the points of 

 divergence, pp. 125129. 



16 Heis, Period. Sternschn. p. 6. (Compare also Aristot. 

 Problem, xxvi. 23 ; Seneca, Nat. Qucest. lib. i. 14 ; " Ventura 

 significat stellarum discurrentium lapsus, et quidem ab ea 

 parte qua erumpit.") I have myself long believed in the 

 influence of the wind upon the direction of the shooting stars, 

 especially during my stay at Marseilles at the time of the 

 Egyptian expeditioa. 



