August 28, 1890] 



NATURE 



417 



another, subducts it, it not being likely that fires should 

 be kindled in the body of that star, and that the matter 

 thereof should at certain times take fire and shine, at 

 other times be extinguist upon the consumption of that 

 matter. ..." 



This, so far as I know, is the first proposed explanation 

 of stellar variability on record. 



The next star in which variability of light was observed 

 was X Cygni. Kirch's observations made in 1686 and 

 subsequent years were communicated to the Royal 

 Society. He observed the star with the aid of an eight- 

 foot tube in August 1687. It became visible to the naked 

 eye in October, increased in brightness and reached a 

 maximum in November, and finally disappeared. This 

 observer also found that the star had always the same 

 brightness at a maximum, and in assigning it a period of 

 404J days, he noted that this duration was irregular. 



These observations bring us to the time of Newton, 

 who at once saw that the cause of true Novas must be 

 distinct from that producing variability pure and simple. 

 He ascribed the sudden appearance of new stars as pos- 

 sibly due to the appulse of comets : — 



" Sic etiam stellie fixae, paulatim expirant in luceni et 

 vapores, cometis in ipsas incidentibus refici possunt, et 

 novo alimento accensce pro stellis tiovis haberi. Hujus 

 generis sunt stellae fixae, quae subito apparent, et sub initio 

 quam maxime splendent, et subinde paulatim evanescunt. 

 Talis fuit Stella in cathedra Cassiopeise quam Cornelius 

 Gemma octavo Novembris 1572 lustrando illam coeli 

 partem nocte serena minime vidit ; at nocte proxima 

 (Novem. 9) vidit fixis omnibus splendidiorem, et luce sua 

 vix cedentem Veneri. Hanc Tycho Brahaeus vidit un- 

 decimo ejusdem mensis ubi maxime splenduit ; et ex eo 

 tempore paulatim decrescentem et spatio mensium sex- 

 decim evanescentem observavit."^ 



But with regard to the ordinary variables, he accepts 

 Bouillaud's suggestion, and adds another : — 



" Sed fixae, quae per vices apparent et evanescunt, quae- 

 que paulatim crescunt, et luce sua fixas tertiae magnitu- 

 dinis vix unquam superant, videntur esse generis alterius, 

 et revolvendo partem lucidam et partem obscuram per 

 vices ostendere. Vapores autem, qui ex sole et steUis 

 fixis et caudis cometarum oriuntur, mcidere possunt per 

 gravitatem suam in atmosphaeras planetarum et ibi con- 

 densari et converti in aquam et spiritus humidos, et 

 subinde per lentum calorem in sales et sulphura et 

 tincturas et limum et lutum et argillam et arenam et 

 lapides et coralla et substantias alias terrestres paulatim 

 migrare." 



Both Montanari in 1669 and Maraldi in 1692 observed 

 that the magnitude of 3 Persei or Algol was variable. 



The information they gave with respect to changes of the 

 star from the second to the fourth magnitude, though im- 

 portant, was not very definite, and it was left to Good- 

 ricke, an English astronomer, to discover, in 1782, the 

 periodicity of these variations and to conclude: — (i) '"That 

 the star changes from about the second to the fourth 

 magnitude in nearly three hours and a half and then 

 back to the second magnitude again in the same time. 

 (2) That this variation occurs about every two days and 

 twenty-one hours." - Flamsteed observed the star in 1696, 

 and found it to be of the third magnitude, and Goodricke, 

 by comparing it with one of his own, deduced the more 

 accurate value of 2 days, 20 hours, 48 minutes, 56 seconds. 

 At the end of the observations Goodricke added the note: 

 — "I should iir.agine that the cause of this variation 

 could hardly be accounted for otherwise than either by 

 the interposition of a large body revolving round Algol, 

 or some kind of motion of its own whereby part of its 

 body covered with spots or such-like matter is periodically 

 turned towards the earth." 



Another variable observed by Goodricke was )3 Lyras. 



' "Principia," p. 525 (Glasgow, 1871). 

 » "Phil. Trans., 1783, p 474. 



NO. 1087, VOL. 42 J 



His first observations brought him to the conclusion that 

 the star had a periodical variation of nearly six days and 

 nine hours, but a further investigation showed that the 

 true period was twelve days nineteen hours, there being 

 two maxima and minima. At one minimum the magnitude 

 of the star is between four and five, at the other between 

 three and four. 



Zollner, in a relatively recent discussion advances very 

 little beyond the views advocated by Newton. In con- 

 sidering the main causes of variability, he lays the greatest 

 stress upon an advanced stage of cooling, and the con- 

 sequent formation of scorias which float about on the 

 molten mass. Those formed at the poles are driven 

 towards the equator by the centrifugal inertia, and by the 

 increasing rapidity of rotation they are compelled to 

 deviate from their course. These facts, and the meeting 

 which takes place between the molten matter, flowing in 

 an opposite direction, influence the form and position of 

 the cold non-luminous matter, and hence vary the rota- 

 tional effects, and therefore the luminous or non-luminous 

 appearance of the body to distant observers. This 

 general theory, however, does not exclude other causes, 

 such as, for instance, the sudden illumination of a star by 

 the heat produced by collision of two dark bodies, 

 variability produced by the revolution of a dark body, or 

 by the passage of the light through nebulous light- 

 absorbing masses. 



Among modern inquirers Prof. Pickering has been 

 more original in his suggestions. He has shown that the 

 light-curves of some stars may be explained by supposing 

 them to have axes of different lengths, with dark portions 

 at the ends, symmetrically situated as regards the longer 

 axis. 



In the following discussion of the cause of variability 

 suggested by the meteoritic hypothesis, I shall divide 

 variability into regular and irregular, defining regularity 

 by constantly recurring maxima and minima on the light- 

 curves. 



The Causes of Variability suggested by the 

 Meteoritic Hypothesis. 



Regular Variability.— P^\ regular variability in the light 

 of cosmical bodies is caused by the revolution of one 

 swarm or body round another (or their common centre of 

 gravity). 



In the case of the revolution of one swarm round 

 another an eUiptic orbit is assumed, and the light at 

 maximum is produced by collisions among the meteorites 

 at periastron. 



In the case of the revolution of a swarm round a con- 

 densed body, the light at maximum is produced by the 

 tidal action set up in the secondary swarm. 



In the case of one condensed body revolving round 

 another, the light at minimum is caused by an eclipse of 

 one body by the other. This can only happen when the 

 plane of revolution of the secondary body passes very 

 nearly through the earth. 



Irregular Variability. — All irregular variability in the 

 light of cosmical bodies is caused {a) by the revolution of 

 more than one swarm or body round another (or their 

 common centre of gravity) ; or (Jb) by the interpenetration 

 of meteoritic sheets or streams. 



In the case of the revolution of more than one swarm 

 round another in elliptic orbits, the irregular maxima are 

 caused by differences of period and periastric conditions 



So far as I know, the only previous explanation of 

 variability on such a basis as the one above stated, which 

 assigns the revolution of one mass round another as a 

 cause of variability, is the one we owe to Newton, who 

 suggested that such stellar variability as we are now con- 

 sidering was due to conflagrations brought about at the 

 maximum by the appulse of comets ; and no doubt his 



