129 CHAPTER IX. VARIABLE STARS OF SHORT PERIOD. WE have seen, in the last chapter, that stars varying their light in periods of less than fifty days stand apart in several important respects from those undergoing slower changes. The distinction is accentuated by the tendency apparent in each class to group its members as far as possible from the frontier-line of separation from the other. Thus, long periods for the most part exceed three hundred days, while a large majority of short periods fall below ten. Thirty-eight stars in all are reckoned as variable within fifty days; of these thirty-two complete an oscillation in less than twenty, twenty-seven (including two with imperfectly ascertained periods) in less than ten days. A comparison of figures 17 and 18 shows that, among short periods taken en masse, those of three to four days predominate; those of five to eight days when Algol variables are excluded. Variables of short period are, as we have said, nearly all white or yellow stars. A very few are reddish; and oneW Virginis—is suspected to possess a banded spectrum. R Lyræ, a star of about 4-5 magnitude, with a superb spectrum of the third type, is nominally variable in forty-six days, but its changes are so trifling and so imperfectly rhythmical as to suggest that its proper place is with ẞ Pegasi and Herculis among stars affected by abortive periodicity. Those characterised by the description of variability we are now studying, seem to be bodies of more finished organisation than long-period variables, which (perhaps for that reason) they largely surpass in light-giving power. It has been remarked1 a 1 Espin, Observatory, vol. v. p. 79. K Numbers of Stars that fluctuations are, on the whole, quick and slight in proportion to the brightness of the objects they affect; and quick fluctuations are executed with much greater precision than slow ones. In many stars the light ebbs and flows like clockwork as to time, and as to measure, with deviations scarcely of the tenth of a magnitude from a settled standard. These remarkable changes progress gradually and continuously in Pickering's fourth class of variables; in his fifth class they Under 14 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 FIG. 17.-Distribution of all the periods of Variable Stars under only interrupt, although at perfectly regular intervals, the usually steadfast shining of certain stars. Of these two kinds, the former is conspicuously exemplified in B Lyræ—a star of which we have already made the acquaintance in connection with its gaseous spectrum-the latter in Algol. The variations of B Lyræ, detected by Goodricke in 1784, were first completely investigated by Argelander in 1844.1 They are of a somewhat complex nature, including two equal Under. 14 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 FIG. 18.-Distribution of periods under twenty days, excluding maxima separated by two unequal minima (see fig. 19). At its highest light the star is of 3-4, at its lowest, of 4·5 magnitude. The intervening minimum is usually at 3.9 magnitude, but at times it tends to become effaced, while at others it is scarcely less marked than the principal minimum. The intervals between all the four phases are approximately equal, but the curve of change is sharpest at the principal minimum, De Stella B Lyra Disquisitio. and it is then, accordingly, that the most exact observations are made. The entire cycle is traversed in twelve days, twenty-one hours, forty-seven minutes, and thirteen seconds, being some two hours and forty minutes more than the time occupied a century ago; and the retardation continues to progress, though by no means uniformly. It is a singular fact that changes in quality are, in this star, combined with changes in quantity of light, yet in a perfectly independent fashion. The bright lines in its spectrum vary seemingly on their own account. Instead of flashing out towards maximum, as in long-period variables, they, if anything, fade, as if overpowered by the general increase of lustre. Mr. Maunder suspected the hydrogen rays to be actually dark, October 19, 1888, when the star was at a Mag. 3·4 3.6 3.8 4:0 4.2 4:4 04 1 2 3 4 5 6 7 8 9 10 11 12 13 FIG. 19.-Light-curve of B Lyra (Argelander). maximum; while on the other hand, the only positive record of their showing by absorption was on September 5, 1882, the day after a secondary minimum. No rule, then, can be clearly made out, and an apparent anomaly has to be admitted. Beta Lyræ is the only short period variable known to give a gaseous spectrum. Its double periodicity is reflected, though in a far less finished form, in R Sagittæ, a star fluctuating comparatively slowly and unpunctually from about the tenth to 8 magnitude. The period, normally of seventy days, shortens and lengthens alternately to the extent of some four days every ten years, and includes two unequal maxima and two unequal minima (see fig. 20). The minima are subject to curious exchanges of intensity. They became equalised in 1873 with 1 'Baxendell, Proc. Phil. Soc. of Manchester, vol. xix. p. 120. 1 the result of cutting the period completely in two; then the phase that had been subordinate grew to be the principal, while the principal declined to a subordinate rank; and there are signs that the original state of things will, after a time, be restored. A pause in the decline, as if a second maximum were Mag. 8.5 9:0 9'5 10'0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 contemplated but failed to be carried out, is a common characteristic of short-period variables. It is well exemplified in Cephei, which has continued, since its discovery by Goodricke in 1784, to oscillate with marvellous regularity between 3.7 and 4.9 magnitudes in a period of five days, eight hours, forty-seven minutes, and forty seconds. A swift ascent is accomplished in 1a 18.6h; then, sixteen hours after maximum, occurs the halt marked by the shoulder of the curve in fig. 21. The spectrum of 8 Cephei resembles that 1 Schönfeld, Vierteljahrsschrift Astr. Ges. Jahrgang xxi. p. 301; Baxendell, Proc. Phil. Soc. of Manchester, vol. xxiv. p. 200. of the sun in being crossed by a great number of fine lines. A wide double star, its full yellow colour contrasts effectively with the delicate blue of a smaller companion, perhaps physically related to, though not visibly circulating round it. Like & Cephei, 7 Aquila never sinks out of reach of the unarmed eye. At its faintest it is of 4-7, at its brightest of 3-5 magnitude. In the mode, no less than in the range of its fluctuations, it closely resembles the Cepheus variable (see its light-curve in fig. 22). Their period of seven days, four hours, fourteen minutes, and four seconds, has probably slightly expanded since Pigott first noticed them in 1784,' and is sometimes irregularly deviated from to the extent of several hours. The spectrum of 7 Aquila is of the solar type. In 10 Sagittæ, a variable discovered by Mr. Gore in 1885, the typical pause in descent is strongly marked. The star Mag. 3.7 3.9 η 4.1 4.3 4.5 Od 1 2 3 4 5 6 7 8 9 10 rises by eight-tenths of a magnitude in less than three days, then drops to its former level in five and a half. In its dim phases it cannot be followed with ordinary sight. The maximum of Geminorum, noticed by Schmidt in 1847 to vary between 3.7 and 4-5 magnitudes, is, by a somewhat rare exception, symmetrically placed as regards the preceding and ensuing minima (see fig. 23). The period of ten days, three hours, forty-three minutes, and twelve seconds has lengthened by ten minutes since its definition forty-three years ago. In one southern star the rule of an ascent quicker than the descent is found to be inverted. This is R Trianguli Australis, discovered by Gould to vary from 6.6 to 8 magnitude, the change upward being accom 1 Phil. Trans. vol. lxxv. p. 127. 2 Sawyer. Astr. Jour. No. 157; Chandler, Astr. Nach. No. 2749. |