too inaccurate to build on. Again, Tycho takes two observations of altitude and azimuth; from the first he computes the declination, corrects this for the motion of the comet in the interval, and with this and the second azimuth computes the altitude for the time of the second observation. For a body as near as the moon there would be a considerable difference, while several examples show none. Finally, Tycho employs the method of Regiomontanus for finding the actual amount of parallax from two altitudes and azimuths, but several combinations gave the same result, that no parallax whatever could be detected in this way. Tycho was well aware that this was a bad method, and evidently only tried it as a duty.[1] (The comet of 1585 was chiefly observed with the large armillæ, and the want of parallax was demonstrated by comparing the right ascension and declination observed with an interval of some hours with the daily motion of the comet.[2])
In the seventh chapter the position of the comet's tail is examined. The increased attention which had been paid to comets during the sixteenth century had led to the discovery of the fact that their tails are turned away from the sun, and not only Peter Apianus, who is generally credited with the discovery, but also Fracastoro, and after them Gemma and Cardan, had pointed out this remarkable fact from observations of different comets. Tycho, who took nothing on trust, examined the matter, and computed from twelve observations of the direction of the tail of the comet of 1577 the position of the tail with regard to a great circle passing through the sun. He found that the direc-
- ↑ See his remarks about the method, De mundi æth. rec. phen., p. 156, and in a letter to Hagecius (who had found a parallax of five degrees by the method), T. B. et doct. vir. Epist., p. 60. Delambre sets forth the method with his usual prolixity in Hist. de l'Astr. du Moyen Age, p. 341; Astr. Moderne, i. p. 212 et seq.
- ↑ Epist. Astron., pp. 16-17.
