the publication quaintly termed "Philosophical Transactions" of the London Royal Society in 1796.
Sir W. Herschel here mentions the number of variable stars, constantly increasing under new discoveries, very naturally predicts that closer observation will be likely to show variability in objects previously unsuspected, and recommends that careful comparisons be made from time to time between neighboring stars all over the heavens, so that any change occurring may be at once detected. The original comparisons accompanying this paper have been of but little use, however; they are interesting chiefly as having been the first attempt to introduce scientific methods into this unexplored territory of the astronomical realm. They were made without the aid of any instrument, and consisted of such indefinite statements as—"Star No. 7 about equal to No. 4, and just perceptibly fainter, or decidedly brighter, than No. 12." The difference of brightness which Herschel considered as "just perceptible" seems to have been from one fifth to one fourth of a magnitude.
That his least appreciable difference should have some constant relation to the traditional "magnitude" was to have been expected, bearing in mind what this oldest and most universal scale of reference was intended to express. The fixed stars were assigned to classes of brightness, we learn, before the Christian era; and the very term "magnitudes," used from the first to designate these classes, shows the state of knowledge under which the study had its origin, for, as we now know, the apparently greater size of the brighter stars is due only to imperfections of the eye. All visible stars—all that existed, that is, for the early astronomers of the Mediterranean—were included in six magnitudes, the first containing the dozen or score of brightest stars in the heavens, the second perhaps twice as many ranking next to these, and so on out in gradually increasing circles. The work of the ancients has in this case been well preserved, no modern innovator having been found bold enough to disturb this time-honored system of reckoning. Still, as in the days of the "Father of Astronomy," the two chief stars of Orion serve as examples of the first magnitude, while his Belt and the Dipper in the northern sky furnish types of the second order. But, while astronomy was yet in its infancy, observers had noticed that the stars were not assorted into well-defined orders, in which all the individuals were equally bright; and so, in assigning to a star its magnitude, they would often add that it was "smaller" or "larger" than the mean of that magnitude. They thus practically trebled the number of their classes. The same division into thirds of a magnitude is still employed by those who judge of brightness by eye estimates, though some are content with dividing into halves, and some undertake to be exact to tenths. Now, even though no scientific precision was attained, or even thought of, in this original apportionment of visible stars among the six magnitudes, to which all later estimates
