METEOES AND AEROLITES. 297 



on these points we owe much to the persevering labours of 

 Brandes and Benzenburg; an ample narrative of whose 

 observations is given in the French work before us. Begun 

 as early as 1798, they were continued at intervals of time 

 and in different places, for a period of thirty-five years. 

 Brandes died in 1834, just after he had received the account 

 of that prodigious fall of shooting-stars in America on the 

 12th and 13th November, which gave at once larger scope 

 and better definition to all our views of these phenomena. 

 To determine the points just mentioned, it was essential to 

 have two observers at least, "and a base of sufficient length 

 for separate observation. Equally essential was it to assure 

 the. identity of the objects seen; for which recourse was had 

 to the exact time of appearance, as well as to the apparent 

 brilliancy, swiftness, and length of train of each star ob- 

 served. Observations strictly simultaneous were needful to 

 success ; and these could only be got by knowing the precise 

 difference of longitude between the stations. The base first 

 taken (two leagues in length) proved too short to furnish 

 the parallax required. In 1801 the enquiry was resumed 

 with the aid of two fresh observers ; and four points were 

 taken, the extremes of which, Hamburgh and Elberfeld, 

 were about 200 miles distant. Here again it may be pre- 

 sumed that the separation was too great, since, out of a great 

 number observed, only five shooting-stars could be actually 

 identified. But this paucity of positive results is familiar 

 to practical astronomy ; and Benzenburg consoled himself in 

 quoting the phrase of Lalande : ' II n'y a que les astronomes 

 qui sachent par combien d'observations manquees on en 

 achete une seule qui reussit.' 



During the remainder of the period we have named, simi- 

 lar observations were repeated by the same and many other 

 observers in various parts of Grermany ; with different lengths 

 of base, and aided by formulae which Olbers and Erman had 



