144 COSMOS. 



Alexandria to 4638,) appears at first sight strikingly small. 5 

 If we assume the moon's mean semi-diameter at 15' 33" -5, it 

 would require 195,291 surfaces of the full moon to cover the 

 whole heavens. If we further assume that the stars are uni- 

 formly distributed, and reckon in round numbers 200000 

 stars from the 1st to the 9th magnitude, we shall have nearly 

 a single star for each full-moon surface. This result ex- 

 plains why, also, at any given latitude, the moon does not 



to the naked eye. An examination of Bode's Uranography 

 (containing 17240 stars), which is composed of the most hete- 

 rogeneous elements, does not give more than 5600 stars from 

 the 1st to the 6th magnitude inclusive, after deducting the 

 nebulous spots and smaller stars as well as those of the 6 -7th 

 magnitude, which have been raised to the 6th. A similar 

 estimate of the stars registered by La Caille between the 

 south pole and the tropic of Capricorn, and varying from the 

 1st to the 6th magnitude, presents for the whole heavens two 

 limits of 3960 and 5900, and thus confirms the mean result 

 already given by yourself. You will perceive that I have en- 

 deavoured to fulfil your wish for a more thorough investigation 

 of these numbers, and I may further observe that M. Heis of 

 Aix-la-Chapelle has for many years been engaged in a very 

 careful revision of my Uranometrie. From the portions of this 

 work already complete, and from the great additions made to 

 it by an observer gifted with keener sight than myself, I find 

 2836 stars from the 1st to the 6th magnitude inclusive for 

 the northern hemisphere, and therefore, on the pre-supposi- 

 tion of equal distribution, 5672 as the number of stars visible 

 throughout the whole firmament to the keenest unaided 

 vision." (From the Manuscripts of Professor Argelander, 

 March, 1850.) 



5 Schubert reckons the number of stars, from the 1st to the 

 6th magnitude, at 7000 for the whole heavens (which closely ap- 

 proximates to the calculation made by myself in Cosmos, p. 140,) 

 and upwards of 5000 for the horizon of Paris. He gives 70000 

 for the whole sphere, including stars of the 9th magnitude. 

 (Astronomie,t}\. iii. s. 54.) These numbers are all much too high. 

 Argelander finds only 58000 from the 1st to the 8th magnitude. 



