4i6 



NATURE 



[August 28, 1890 



the result of the orderly condensation of a single swarm, 

 for the simple reason that we shall have two swarms or 

 bodies to deal with. Even if the very highest tempera- 

 ture is reached, we shall not have exactly the same spec- 

 trum as that presented by Group IV. 



The most stupendous case illustrating the above re- 

 marks is to be found in the Pleiades, the true structure of 

 which has been revealed to us by Mr. Roberts. The 

 principal stars are not really stars at all ; they are simply 

 loci of intercrossings of meteoritic streams, the velocities 

 of which have been sufficiently great to give us, as the 

 result of collisions, a temperature approaching that of a 

 Lyrae, so far as we can judge by the spectrum ; but that 

 the a Lyrae conditions are not present is evidenced by the 

 fact that in Pleione the broad dark hydrogen lines have a 

 bright thin line running down their centres, indicating 

 that we have intensely-heated hydrogen outside that 

 which is absorbing. 



So long as these meteoritic streams are interpenetrating 

 and disturbing each other, so long the Pleiades will shine ; 

 but their light may soon cease if the disturbance comes 

 to an end, for we are not dealing with masses of vapour 

 like a Lyrae. Indeed, one of them seems to have already 

 become invisible. Of the seven daughters of Atlas, one has 

 disappeared. The " septem radiantia sidera " are seven 

 no longer. The seventh had vanished before the time of 

 Aratus. 



" The Pleiades ; small the region 



They fill, and pale the light they dart. 



Seven journeyers men call them 



Though only six are visible to ken. 



No star, I wis, has vanished from Heaven's floor 



Within mortal tradition, and idly is that number 



Fabled. Natheless seven the names they bear : 



Alcyone, Merope, Cela;no, tlectra, 



Sterope, Taygete, and stately Maia." ' 



At the beginning of the action to which I have ascribed 

 the present light of the Pleiades, we should have the 

 appearance of a "new star," and the greater the light 

 produced and the more sudden the outburst the more 

 certainly would the appearance of a new star be chronicled. 

 Many such stars have burst forth, and the phenomena re- 

 corded have been entirely in harmony with the explana- 

 tion afforded by the hypothesis ; but, as the discussion 

 of these phenomena is not yet complete, I shall not in 

 the present article touch further upon them ; but I may 

 point out that, before the existence of " variable stars " 

 was recognized, as it is now, the increase in magnitude of 

 a variable at maximum, rendering visible to the naked 

 eye what was before invisible, was attributed to the 

 creation of a new star. Hence it is that the first work 

 done on the periodicity of variable stars grew out of 

 observations of so-called Novae. 



Leaving on one side, then, any question of Novae, we will 

 inquire into the growth of our knowlege of stars the light 

 of which is known to wax and wane with more or less 

 regularity, and see to what causes this variability has 

 been ascribed. We have to consider those shorter periods 

 of light-variation, well within human ken, light-changes 

 which, instead of taking milHons and perhaps billions of 

 years, are undergone in a few days, or weeks, or months. 

 Such changes have been abundantly chronicled from the 

 earliest times and acknowledged to be among the most 

 mysterious phenomena presented to us in the skies. 



In this historical survey we must first consider the case 

 of Mira or o Ceti. It is now nearly three centuries ago 

 since Fabricius noticed this star (August 1596), thinking 

 it to be a nova, and watched its disappearance in the 

 following October.^ 



Not only Fabricius but Kepler looked upon Mira Ceti 

 as a new star similar to those of 1572 and 1604. Indeed, 

 it was regarded as such until 1638, when some observa- 

 tions by Phocylides Holwarda brought out for the first 

 time the fact that the changes in magnitude repeated 



' Poste's translation, p. 13. 



" Kepler, "De Stella," chap. xii. p. 112. 



NO. 1087, VOL. 42] 



themselves. The work done by this astronomer is Quoted 

 by Hevelius. 



Holwarda first observed the star in December 1638,. 

 when it was brighter than a third magnitude ; he watched 

 it decrease to the fourth, and disappear during the sum- 

 mer of 1639. On December of the same year he again 

 observed it. There is no doubt, indeed, that Holwarda 

 was the first to demonstrate by these observations that 

 the light of stars is liable to periodic changes in intensity. 



FuJlenius, a teacher of mathematics at Franecker, was 

 the next to observe Mira. He noted that the star was 

 visible on September 23, 1641, and the same date in the 

 following year. In August 1644, however, no trace of it 

 could be made out. 



Junquis, a professor at Hamburg, recorded that Mira 

 was of the third magnitude on February 18, 1647, and 

 was invisible from July 1648 to November of the same 

 year. 



It was Hevelius, however, who made the first detailed 

 investigation into the variations of the light of this star. 

 Beginning in January 1648 he assiduously watched the 

 changes in magnitude until March 1662, and placed the 

 question of variability beyond the possibility of a doubt. 



During the fifteen years of observation Hevelius saw 

 the star go through its changes in magnitude many 

 times, and noted that it was always invisible for several 

 months in the year. He did not, however, determine 

 the period, although it will be seen that the following 

 observations would have been sufficient for him to have 

 deduced an approximate value : — 



Sept. 10, 1660 — "Instar stellae 4 magn. fere." 

 Aug. 20, 1661 — " Vix quartse magnitudinis extitit." 



Inierval, 344 days. 



Sept. 20, 1660— "^qualis illi in ore Ceti." 



Aug. 29, i66i — "Squalls illi in ore Ceti." 



Interval, 353 days. 



The determination of the period of Mira Ceti was 

 deduced by Bouillaud in 1667 from all the observations 

 which had been made from its discovery in 1638 to 1660. 

 This discussion occurs in a rare book having the title 

 " Ismaelis BuUialdi ad Astronomos monita duo : Primum 

 de Stella Nova, quae in Collo Ceti ante annos aliquot 

 visa est. Alterum, de nebulosa in Andromedas cinguli 

 parte Borea, ante biennium iterum orta." 



A review of the book appeared in the first volume of 

 the Philosophical Transactions (p. 381), from which the 

 following account of Bouillaud's conclusions have been 

 taken ; — " . . . That one period from the greatest phase 

 to the next consists of 333 days ; but that the interval of 

 time betwixt the times of its beginning to appear equal 

 to stars of the sixth magnitude, and of its ending to do 

 so consists of about 120 days; and that its greatest 

 appearance lasts about fifteen days : all which yet he 

 would have understood with some latitude. 



"This done, he proceeds to the investigation of the 

 causes of the vicissitudes in the emersion and disap- 

 pearance of this star, and having determined that the 

 apparent increase and decrement of every lucid body 

 proceeds either from its changed distance from the 

 eye of the observer, or from its various site and position 

 in respect of him, whereby the angle of vision is changed, 

 or from the increase or diminution of the bulk of the 

 lucid body itself; and having also demonstrated it im- 

 possible that this star should move in a circle or in an 

 ellipsis, and proved it improbable that it should move in 

 a strait line, he concludes that there can be no other 

 genuine, or at least no other more probable cause of the 

 emersion and occultation than this : That the bigger part 

 of that round body is obscure and inconspicuous to us, 

 and its lesser part lucid, the whole body turning about its 

 own center and one axe, whereby for one determinate 

 space of time it exhibits its lucid part to the Earth,^ foi 



^ Here we have the germ of Sir Wm. Herschel's reference to the action oi 

 varying amounts of spotted surface ; Maupertius' idea of rotatory disks ; and 

 Prof Pickering's suggestion of axes of different lengths. 



