Very little stress can ever be laid on the personal appearance of 2 comets because, whilst many of them resemble one another very closely, the same comet at different epochs has often been known to present very different appearances. Reliance must not in all cases be placed on an apparent similarity of elements even where similarity to a striking degree seems to exist.[1]
When a new comet has been found it is a matter of the greatest importance to determine very accurately its position in the heavens from day to day; and this is sometimes not very easy, especially when the comet is viewed in twilight. But whatever may be its place, that is determined by measuring, by means of a micrometer, its angular distance from particular stars, whose exact position in Right Ascension and Declination is either accurately known, or can be ascertained at leisure. The stars used for this purpose are spoken of as "comparison stars". Ordinarily an equatorial has to be made use of, and its circles should of course be in very accurate adjustment. If, however, by good fortune the comet can be caught on the meridian and seen through a meridian instrument the resulting places will usually be more accurate than if an equatorial is employed. Only 3 perfect observations are necessary for determining the general nature of a comet's orbit; but at best the first result will only be provisional, especially if the intervals between the observations are short, such as 2 or 3 days. Such observations will only yield an orbit in the form of a provisional parabola. If by any chance the comet is moving in an elliptic orbit, the intervals must amount to 2 or 3 weeks at the least for the character of the ellipse to be ascertained with any reasonable accuracy. The plane of the orbit and the comet's perihelion distance, ascertained provisionally, will not generally be varied much by the utilisation of subsequent observations; but it is another matter to determine accurately the eccentricity of the orbit, the dimensions of the major axis, and the corresponding period. The more the observations are prolonged the more the figures for these three elements will vary from those first
- ↑ See p. 18 (ante).
