OCTOBEN. 1011. 



KNOWLEDGE. 



383 



Capture Theory " in order to emphasize this same 

 fact : he has independently deduced the idea that 



of a httle over seven days period. We must 

 assume that in so close a pair, tidal action must ha\e 



Time. 



Figure 13. Deduced light curve of single variable in which 



the impact was of recent date, hence the liglit rises iiuickly 



and falls slowly. 



.^nollier varialile is probably near thi^ star, or it may be one of an orbital pair : its 

 spectrogram should show bright hydrogen and perhaps other bright lines during 



portions of each revolution : nebulosity at minimum may be expected. 

 The spectral lines of planetary nebula should be looked for in all spectroscopic 

 binaries, all variable and Wolf-Rayet stars. The pairs of variables on page Sg, 

 " Rirth of Worlds and Systems," should be examined both for planetary nebulae lines 

 :iud for nebulosity. Midway between the pairs, the meteoric residue of the third 

 body should he sought for. 



the planets could not have ori^dnated from the Sun. 

 and has most conclusively demonstrated the erroneous 

 nature of all the old hypotheses. Thirt\' vears a,i,'o I 

 showed that the surface velocity of the Sun woiikl 

 have to he increased some fort\- thousand times its 

 present amount to enable matter to separate itself 

 and remain indei)endent of the Inuh' of the Sun. 



Since I have been in 

 England I have devoted 

 man\' months to the 

 study of the light curves, 

 the spectrograms, and 

 other peculiarities of vari- 

 able stars, and there is 

 not one amongst the whole 

 that I have studied but 

 seems to have originated 

 in some form of impact. 

 The extremely singular 

 light curves and eccentric 

 distribution of variables 

 in star clusters received 

 fairly satisfactory explana- 

 tions. The eccentric light 



Time. 



Time. 



Figure 14. Deduced light curve of a non - eclipsing 

 spectroscopic binary. 



Associated by impact at a comparatively recent date. 



Tidal action has made the time of rotation of each sun synchronous with the period 



of revolution. 

 The volcanoes remain active in each sun. The phases of the volcanoes being 

 approximately half a revolution apart, the spectrosrams should intermittently show 



bright hydrogen lines. 

 Nebulosity shi-mld be looked for with long exposure and great light gathering 



power. 



made rotation isorhmnous with re\'olution. It 

 appears that the impact occurred a sufficiently short 

 time ago that the N'olcanoes of impact exist in both 

 the stars. .Vs the pair re\'oh'e, first one of these and 

 then the other comes into view, giving us the well- 

 known curve of the Cepheids. Perhaps the most 

 wonderful example of the combination of many 



deductions in one pair is 

 the doubly \- a r i a b 1 e 

 nebulous double star, with 

 an oscillatory orbit, that 

 we worked out in Sydney 

 during the time I was 

 there t\\ o \ears ago. This 

 star, also S.S. Signi, and 

 a number of other peculiar 

 \-ariables, I have described 

 in detail in '" The Birth of 

 Worlds and Systems." 



In the next article, after 

 a few words devoted to 



Figure 15. Deduced light curve of eclipsing spectroscopic star clusters and planetary 



binary produced by impact of nearly a third, hence the pair nebulae, we will discuss 



has just escaped whirling coalescence. jj^g phenomena that may 



curves of the Cepheids The cuts in the tom suns were so deep, and the third body consequently so l.arge. be deduCed frOm the 



, . r that each star has been greatly heated and so extensively bathed in the volatilized ■ f. f \\rU,t„ NAVinlof. 



received most SatlSiaCtOrV shininglluid, .as to have become a luminous sun. Much of the third star will rem.aiu lUipaCL Ul Willie INCUUlclc 



, . ^ . ., ■■ partly as a tongue of fire joining the two torn suns ; partly .as a nebulous and gaseous j j.u intemenetration 



explanation. Eta AqUllae .atmosphere this Umer giving bright liues. Perchance the nebulosity may occasion- dUU lUe lU lei peiieu atlOll 



ic r^f.rVianc the- most ally give the star a h.azy delinition, especially with long exposure. In case the ^^ y^^^ Stellar SvStCmS 

 IS pel uapb Ulie IIIUM i,„p,^(.t that produced this combination be of recent date, a hump of light may 



t\-i-,if~nl rvf tlipcp cfni-Q Tt indicate the volcano of impact on one or both the suns, and may show on the light sUCll 



INpicai Ul llicsc SLclt^. It ,^^^^^^ Possiblv the lines of a pLanetary nebula may be detected if carefully sought ^^ 



is a SpeCtrOSCOI>ic binar\' for by means of an analysing spectroscope. "" 



as the Magellanic 



Clouds or the Galax\- itself. 



jvIoTK. — Association of \'aki.\hlk St.vks with Nebuloi:s M.\tter. — It is well to understand 

 the thermodynamic intensit\- of the volcanic ejection of solar volcanoes. Ejections from so calm a body 

 as our Sun 'run into velocities of hundreds of miles a second. Compare the speed of 200 miles with 

 half-a-mile a second, which is the velocity of our swift projectiles. (200 X 2)'^= 160,000, that is, the energy 

 of a Krupp shell would be 160,000 times greater than it now is, were it moving as s\\ifth- as some solar 

 protuberances. This then is the order of the Kinetol of projection in the torn suns. 



