Mabch 1, 1892.] 



KNOWLEDGE 



51 



earth and Moon to have their present diameters and 

 to be homogeneous bodies* when the Moon was at a 

 distance of 10.000 miles from the earth's centre. 



Until Roche's limit was approached the summits of the 

 terrestrial and lunar atmospheres would be widel,y separated 

 by a gulf of more than rSOOO miles, and unless one or both 

 bodies were very hot, and the heights of their atmospheres 

 were greatly raised by temperature, we cannot suppose 

 that there would be any intermixture of the atmos- 

 pheres and consequent drafting of gas from one body to 

 •the other. If, however, they were both in a sun-like 

 condition and their atmospheres or coronas intermixed, 

 we should have an exchange of gaseous matter going on. 



If they were both about the same temperature, the 

 smaller body having the least gravitating power, would 

 have the highest atmosphere, and a transfer of gaseous 

 matter would go on from the smaller body to the larger 

 body. If, however, the smaller body cooled first and the 

 larger remained in the sun-like condition, we might have a 

 transfer of gaseous matter from the larger to the smaller 

 body, owing to a condensation of vapour on the side of 

 the smaller body remote from the larger. Such exchanges 

 may possibly account for the different density of the earth 

 and the iloou,' and also, as I have suggested in another 

 place, for the different colours , of binary stars, and the 

 curious fact that the smaller star of a binary is apparently 

 always bluer than the larger star. 



STELLAR SPKCTRA. 

 To the Editor of Knowledge. 

 Sir, — In Knowledge for June, 1891, Mr. Maimder 

 annoimced, for I believe the first time, a principle as to 

 the relative brightness of stars with different spectra, 

 which seems destined to bear very important fruits in 

 Astronomy. I have lately compared the stars treated of 

 by him with the Draper Catalogue of the Harvard 

 University, with the result of removing most of the 

 anomalies in his table, save that of the extraordinary 

 brilliaucy of -/ Leonis. Four stars with spectra of the 

 first type (Sirians) in Mr. Maunder's list, have a relative 

 brightness of less than iO. These are O S 4, relative 

 brightness 1--18, J Caucri, relative brightness 2-90, 

 i; Coronte BoreaUs, relative brightness 1-40, and //.- Bootis, 

 relative brightness 2-74. Mr. Gore (whose figures Mr. 

 Maunder follows) seems to have imder-rated the magnitude 

 of the first of these ^tars. Its photographic magnitude is 

 6-46 according to the Draper Catalogue, if I am right in 

 my identification, and I find that even with Sirian stars the 

 photographic magnitude is usually less than the photometric. 

 Taking its magnitude at 6-46 its relative brightness is 

 ;r53 instead of 1-48. Otherwise it is not in the Cata- 

 logue, g Cancri, according to the Draper Catalogue, has a 

 spectrum of the Solar not the Sirian type, and the 

 same remark applies to tj Coron.'P Borealis. a- Bootis 

 is not in the catalogue at all. The star with a spectrum 

 of the first type is p Bootis, a brighter star which is situated 

 near the binary pair, and I suspect that it was the spectrum 

 of this star that Mr. tiorc inserted in his table. 4-00 thus 

 seems to be the minimum of relative brightness for a Sirian 

 star, if indeed any of them sink so low, while there are only 

 five solar stars in Mr. Maunder's table (besides y Leonis) 

 which rise so high. The brightest of these is it Cephei 

 11"07, the spectrum of which is marked with a ? in the 

 Draper Catalogue, and from a note it appears that on 

 special examination the spectrum was considered to be of 

 the thiiil type. That stars of this type should be relatively 



* The earth's deiisitv being taken as unitv the Jloon's density 

 is 0-613. 



brighter than those of the second type (Solar stars) is no 

 doubt an unexpected result, but it is confirmed by the next 

 case, 36 Andromed* 0-28, whose spectrum is. also referred 

 to the third type in the Draper Catalogue, r Ophiuchi 

 7"35 is the brightest of the remainder, Doberck's orbit beiug 

 adopted. As difl'erent computers have difi'e red considerably 

 with regard to the period of this star, I think the orbit 

 would be worth re-computing, including recent measure- 

 ments. The relative brightness of the remaining two is 

 under o-QO. I may note that several stars classed by Mr. 

 Gore and Mr. Maunder as Sirian, appear as Solar stars 

 (spectrum F or E ) in the Draper Catalogue. Their spectra 

 seem to lie on the border-land, and Pickering gives a wider 

 extension to the second type of spectra than to the first. 

 But these stars confirm the general conclusion drawn by 

 Mr. Maunder, for their relative brightness is in all cases 

 below the Sirian average and above the Solar average. 

 Their intermediate position is thus established. 



I remain, your obedient Servant, 

 January 6th, 1892. W. H. S. Monck. 



P.S. — I find that I was in error in stating that the 

 photographic magnitudes are usually less than the photo- 

 metric in the case of stars with spectra of the first type. 

 With the brighter stars they are almost invariably less, 

 but with the fainter stars they are often greater. It is, 

 in fact, clear that the two scales do not proceed on a 

 common basis. ' W. H. S. M. 



THE IXHEREXT LIGHT OF JUPITER. 

 To the Eilitor of Knowxedge. 



Sir, — Among the " other considerations " referred to by 

 i\Ir. Gore in his very interesting article (on page 3G), whence 

 the self-luminosity of Jupiter may be inferred, perhaps I may 

 be allowed to recall those derivable from an observation of 

 mv own made in 1880, of which the details will be found 

 on page 47 of vol. XLI. of the R.A.S. Monthlij y<diees. On 

 the occasion to which I refer the shadow of Satellite II. 

 was distinctly of a chocolate colour ; and the only feasible 

 explanation that I could find of this phenomenon was, that 

 the portion of the planet upon which the shadow was pro- 

 jected must have been red-hot. If we were to illuminate a 

 whitewashed wall by light from a lamp or lantern trans- 

 mitted through a red glass, obviously tbe wall would 

 appear red. If now, we Hooded it with the electric 

 hght, this would wholly overpower the more feeble illumi- 

 nation, and the wall would appear white. It is evident, 

 though, that any object interposed between the electric 

 light" and the wall would cut off the more brilliant form of 

 illumination, leaving the feebler one to assert itself ; and 

 that hence shadows would appear not black but red. I 

 may add that Mr. Campbell, F.R.A.S., observed this brown 

 shadow of Jupiter's second satellite at Barnet, simul- 

 taneously with my own observation of it in Sussex. 



I am, faithfully yours. 

 Forest Lodge, ]\Iaresfield, Susses, Willlwi Noble 



February 2nd, 1892. 



[The intensity of the inherent light of Jupiter must be 

 small as compared with the brightness of the sun's 

 light at the distance of Jupiter, for the satellites are 

 entirely lost to our view when they plunge into the shadow 

 of Jupiter. If their brightness were reduced only eight 

 stellar magnitudes when they plunged into the shadow of 

 the planet they would still easily be visible in large tele- 

 scopes. For the third satellite, which is the brightest, is 

 usually estimated as equivalent in brightness to a star of 

 the sixth magnitude. Therefore, we may assert that the 

 illumination derived from the whole disc of Jupiter (which 

 subtends an angle of nearly 19°, as seen from the nearest 



