Ptolemy and all the ancient writers described the constellation figures as they are seen on globes, that is from the outside. Bayer in his Atlas, published in 1603, reversed the figures to show them as they would be seen from the _interior_ of a hollow globe and as, of course, they are seen in the sky. Hevelius again reversed Bayer's figures to make them correspond with those of Ptolemy. According to Bayer's arrangement, Betelgeuse (α Orionis) would be on the left shoulder of Orion, instead of the right shoulder according to Ptolemy and Al-Sufi, and Rigel (β Orionis) on the right foot (Bayer) instead of the left foot (Ptolemy). This change of position has led to some confusion; but at present the positions of the stars are indicated by their Right Ascensions and Declinations, without any reference to their positions in the ancient figures. The classical constellations of Hipparchus and Ptolemy number forty-eight, and this is the number described by Al-Sufi in his "Description of the Fixed Stars" written in the tenth century A.D. Firminicus gives the names of several constellations not mentioned by Ptolemy. M. Fréret thought that these were derived from the Egyptian sphere of Petosiris. Of these a Fox was placed north of the Scorpion; a constellation called Cynocephalus near the southern constellation of the Altar (Ara); and to the north of Pisces was placed a Stag. But all these have long since been discarded. Curious to say neither the Dragon nor Cepheus appears on the old Egyptian sphere.[389] Other small constellations have also been formed by various astronomers from time to time, but these have disappeared from our modern star maps. The total number of constellations now recognized in both hemispheres amounts to eighty-four. The first catalogue formed was nominally that of Eudoxus in the fourth century B.C. (about 370 B.C.). But this can hardly be dignified by the name of catalogue, as it contained only forty-seven stars, and it omits several of the brighter stars, notably Sirius! The first complete (or nearly complete) catalogue of stars visible to the naked eye was that of Hipparchus about 129 B.C. Ptolemy informs us that it was the sudden appearance of a bright new or "temporary star" in the year 134 B.C. in the constellation Scorpio which led Hipparchus to form his catalogue, and there seems to be no reason to doubt the accuracy of this statement, as the appearance of this star is recorded in the Chinese Annals. The Catalogue of Hipparchus contains only 1080 stars; but as many more are visible to the naked eye, Hipparchus must have omitted those which are not immediately connected with the old constellation figures of men and animals.