Nov. 1, 1885.] 



KNOWLEDGE 



17 



let it in through a verj' narrow slit, and place the refrac- 

 ting edge of our prism parallel to the length of such 

 slit, we shall obtain a very mueh purer Sj.ectrum, in fact 

 one in which the overlapping of the colours is reduced 

 to a minimum ; and any particular tint separated as above 

 described from such a spectrum, is quite indecomposable 

 by refraction through a second prism placed behind the 

 aperture s in the screen. As might, of coui'se, have been 

 anticipated, if we receive uur spread-out spectrum on a 

 concave mir/or, so as to recombine the various colours, 

 the resulting image projected on to the ceiling or a screen, 

 will b:' a spot of white light. There i.s this notable 



Fig. 3. 



difference between the pure colours furnished by the 

 prismatic image, and those emanating from objects lighted 

 by ordinary sunlight : that the latter are always paler 

 than the colours of the spectrum, owing to the admixture 

 of white light with them ; the green of a leaf for instance 

 falls short of the green of the spectrum in intensity. 

 It is diluted in this way. And even the strange and 

 almost bizarre colours produced by polarised light fall 



A! 



B 



^ 



Fig. 4. 



slightly short of the purity of those of the prismatic 

 spectrum, as may be seen by splitting off a very thin 

 film of selenite, and inserting it through the slit S in 

 the canister in Fig. -t (here reproduced from p. 269, 

 Vol. VIII.), so that the beam of light from the bundle 

 of mirrors passes through it tn route to the smaller 

 one in the lid. A magnifying glass may be used 

 for this observation, if its focus be long enough to 

 enable it to be held at a sufficient distance from the 

 little mirror. Tartaric acid, too, crystallised on a plate 

 of glass and placed in a similar position, exhibits gorgeous 

 colours ; yet, as we have just remarked, not quite equal 

 in vividness and purity to those which we obtained by 

 the aid of our prism, an observation equally applicable 

 to the rays of colour in a .soap-bubble about to burst. 

 {To be continuid.) 



OUR GALAXY. 



letter to sir j. herschkl. 



By Rr'harh a. Pructob. 



London, July 24, 1869. 



HAVE for several months had it in my 

 thoughts to write to you respecting certain 

 views which seem to me to result from what 

 has been discovered respecting stars and 

 nebula', and in particular from the unrivalled 

 series of observations which the scientific 

 worldowes to your father and youi'self. I have, 

 however, had several reasons for delaying. I have not 

 cared to bring the matter before you until I had gathered 

 a sufficient amount of evidence — and also I have had some 

 diffidence in making a statement which sounds at a fir.st 

 hearing as though intended for a correction of the im- 

 pressive theories of Sir W. Herschel ; though all who 

 have really studied his successive essays on the universe 

 will recognise in my work merely an attempt to advance 

 on the road which he pioneered. If the attempt is a 

 failure you will not the less readily recognise its object. 



Although the evidence I have now gathered is far from 

 being so complete as could be wished, yet it is I think 

 sufficient to support many important general conclusions. 

 If it might not seem so to others I think it will to yo\i, 

 who have the means of forming a much readier decision 

 than any man of science now living could do. This is, 

 in fact, one of my reasons for submitting my views to 

 you rather than to any other. There are many eminent 

 mathematicians, many profound physicist.s, but I know 

 of no one but yourself who is at once mathem-tician, 

 physicist, and observer, while possessing besides, that 

 knowledge of the whole range of astronomical facts 

 which seems required in deciding on matters such as 

 those I wish to submit to you. 



My object is, briefly-, to show that there are reasons for 

 modifying our views as to the distribution of stars and 

 nebulii:' throughout space. Our own galaxy appears to 

 be rich in a variety of forms of matter or of aggregation 

 not hitherto ascribed to it. It would seem that within 

 a single region of our sidereal system there may be col- 

 lected, — stars of every variety of magnitude, star-groups of 

 every degree of resolvability, star-streams, gaseous. masses, 

 gaseous systems, and in fact all the features which we 

 have been in the habit of looking upon as characteristic 

 (not of any finite poi-tion of our system but) of that 

 quasi-infinite expanse of discernible space whereof our 

 own galaxy has been thought to occupy but a corner. 



I will present the points of the evidence as they occur 

 to me, without in general indicating their bearing on my 

 views (to save time and also because yovi will see in a 

 moment how each part of the evidence bears on the case). 

 Manj- of the facts are of course already well known to 

 you : but I mention them to make the case complete. 



1°. When the irresolvable nebula? in your large cata- 

 logue are isographically distribiited (in the manner 

 carried out by you on a smaller number) we find that 

 they show a marked preference for extra-galactic space. 

 Their withdrawal from a given great-circular zone, if not 

 accidental, indicates their association with the sidereal 

 scheme by some law. And, as the separate nebul» have 

 every variety of position, the coincidence of our own 

 galaxy (ussumed to be one of the nebula^) with the vacant 

 region — both in place and in position or direction — is too 

 remarkable to be readily accepted as but an accident. 



2°. The very easily resolvable nebulse affect the galactic 

 zone very decidedly, the less easily resolvable nebulae 



