120 COSMOS. 



frequent reference is made to the number of the largest and 

 most luminous or of the dark and less easily recognized stars ; 61 

 but we find no relative comparison of the stars contained in 

 the different constellations. The Catasterisms are, according to 

 Bernhardy, Baehr, and Letronne, more than two hundred years 

 less ancient than the catalogue of Hipparchus, and are besides 

 a careless compilation and a mere extract from the Poeticum 

 Astronomicum (ascribed to Julius Hyginus), if not from the poem 

 Epp.rjs of the older Eratosthenes. The catalogue of Hipparchus, 

 which we possess in the form given to it in the Almagest, 

 contains the earliest and most important determination of 

 classes of magnitude (gradations of brightness) of 1022 stars, 

 and therefore of about ih of all the stars in the firmament 

 visible to the naked eye, and ranging from the 1st to the 6th 

 magnitude inclusive. It remains undetermined whether these 

 estimates are all due to Hipparchus, or whether they do not 

 rather appertain in part to the observations of Timocharis or 

 Aristyllus, which Hipparchus frequently used. 



This work constituted the important basis on which was 

 established the science of the Arabs and of the astronomers 

 of the middle ages : the practice, transmitted to the nine- 

 teenth century, of limiting the number of stars of the first 

 magnitude to 15 (although Madler counts 18, and Riimker 

 after a more careful observation of the southern celestial 

 hemisphere upwards of 20) takes its origin from the classifi- 

 cation of the Almagest, as given at the close of the table of 

 stars in the eighth book. Ptolemy, referring to natural vision, 

 called all stars dark which were fainter than those of his 6th 

 class ; and of this class, he singularly enough only instances 



61 Eratosthenes, Catasterismi, ed. Schaubach, 1795, and 

 Eratosthenica, ed. G. Bernhardy, 1822, p. 110-116. A 

 distinction is made between stars \ap,7rpovs (/zeyaXous-) and 

 apavpovs (cap. 2, 11, 41). Ptolemy also limits of dfj.6p<pc0Tot 

 to those stars which do not regularly belong to a constellation. 



