With reference to the Greek letters given to the brighter stars by Bayer (in his Atlas published in 1603), and now generally used by astronomers, Mr. Lynn has shown that although "Bayer did uniformly designate the brightest stars in each constellation by the letter α,"[454] it is a mistake to suppose--as has often been stated in popular books on astronomy--that he added the other Greek letters _in order of brightness_. That this is an error clearly appears from Bayer's own "Explicatio" to his Atlas, and was long since pointed out by Argelander (1832), and by Dr. Gould in his _Uranometria Argentina_. Gould says, "For the stars of each order, the sequence of the letters in no manner represents that of their brightness, but depended upon the positions of the stars in the figure, beginning usually at the head, and following its course until all the stars of that order of magnitude were exhausted." Mr. Lynn says, "Perhaps one of the most remarkable instances in which the lettering is seen at a glance not to follow the order of the letters is that of the three brightest stars in Aquila [Al-Sufi's 'three famous stars'], γ being evidently brighter than β. But there is no occasion to conjecture from this that any change of relative brightness has taken place. Bayer reckoned both of these two of the third magnitude, and appears to have arranged β before γ, according to his usual custom, simply because β is in the neck of the supposed eagle, and γ at the root of one of the wings."[455] Another good example is found in the stars of the "Plough," in which the stars are evidently arranged in the order of the figure and not in the order of relative brightness. In fact, Bayer is no guide at all with reference to star magnitudes. How different Al-Sufi was in this respect! The stars Aldebaran, Regulus, Antares, and Fomalhaut were called royal stars by the ancients. The reason of this was that they lie roughly about 90° apart, that is 6 hours of Right Ascension. So, if through the north and south poles of the heavens and each of these stars we draw great circles of the sphere, these circles will divide the sphere into four nearly equal parts, and the ancients supposed that each of these stars ruled over a quarter of the sphere, an idea probably connected with astrology. As the position of Aldebaran is R.A. 4{h} 30{m}, Declination North 16° 19', and that of Antares is R.A. 16{h} 15{m}, Declination South 25° 2', these two stars lie at nearly opposite points of the celestial sphere. From this it follows that our sun seen from Aldebaran would lie not very far from Antares, and seen from Antares it would appear not far from Aldebaran. The following may be considered as representative stars of different magnitudes. For those of first magnitude and fainter I have only given those for which all the best observers in ancient and modern times agree, and which have been confirmed by modern photometric measures. The Harvard measures are given:--