Aug. 17, 18S3.] 



» KNOV\^LEDGE ♦ 



107 



absorptive action of the atmospheric medium, yet on 

 the other hand the planet's surface is so much the 

 more brightly illuminated, and the light so illuminating 

 it passes so much the more readily on its return 

 journey through the planet's atmosphere. On the other 

 hand, if a planet's atmosphere is dense, although such an 

 atmosphere much more completely absorbs the light which 

 enters it, yet, on the other hand, it resists the entrance (so 

 to speak) of a much greater proportion, and moreover it is 

 itself illuminated by that which it absorbs. The compen- 

 sation thus effected may not be, and probably is not coqi- 

 plete ; but we are precluded (at least until we have much 

 more exact information than has yet been obtained) from 

 discussing satisfactorily the effects due to absorption by the 

 planet's atmospheres. A similar remark applies to the 

 difference between the absorptive effects of a planet's 

 atmosphere near the centre and near the edge of the disc. 

 At the centre more light passes through the atmosphere, 

 and so the planet's surface is more brilliantly illuminated 

 (apart from the consideration of the angle of incidence). 

 At the edge the light has a longer course through the 

 atmosphere, and is therefore much more freely absorbed, 

 so that the planet's surface is much less fully illuminated 

 than it would he if there were no atmosphere. Neverthe- 

 less, there is no corresponding falling ofi' in the apparent 

 brilliancy near the edge, for the atmosphere is itself illumi- 

 nated. Whether this illumination will be greater or less 

 than that which would result if there were no atmosphere, 

 will depend on the nature, extent, and condition of the 

 atmosphere. 



Passing from such considerations as these, which are of 

 too difficult and complicated a nature to be here satisfac- 

 torily discussed, let us consider how far our estimates of 

 relative brightness are to be depended upon. It must be 

 remembered that in all observations we have to deal with 

 the experience of the senses, and that the senses are 

 fallible. It is important that we should know to what 

 extent the senses are fallible in this particular matter of 

 the estimation of light. 



Unfortunately, we find the result of such an inquiry not 

 altogether encouraging. There is, perhaps, no subject of 

 research in which the senses are more apt to be deceived 

 than they are in dealing with different degrees of lumi- 

 nosity. 



For instance, I have often been asked if it is possible 

 to believe that the faint light of " the old moon in the 

 new moon's arms" is equal to that due to full moon- 

 light ; still less, it is urged, can it bo equal to that due to 

 full earthlight. Yet the darker part of the moon's disc at 

 the time of new moon is illuminated by full, or nearly full, 

 earthlight. Hence, some imagine that distance must 

 reduce the apparent brilliancy of the light received from 

 that portion of the moon. Of course I do not refer to this 

 mistake for the sake of refuting it, but it affords an excel- 

 lent illustration of the way in which the senses may be 

 deceived. Let any person look at a distant hill 

 bathed in full moonlight, note carefully its apparent 

 brilliancy, and endeavour to retain the remembrance 

 of it. Sixteen or seventeen days or so afterwards 

 let him observe the young moon, and note tiie faint 

 light of the part which is not in sunlight, and compare that 

 light with till! light of the moon-illumined hill he had 

 formerly observed. It will appear manifest, and so far as 

 the sens(> of sight is concerned it is msuiifest, that the light 

 of the hill is far the brighter. Nevertheless, nothing can 

 be mor(i c(>rtain than that the light of the old moon is ten 

 or twelve times as bright. The explanation is simple. 

 When we are looking at ttie new moon, we see the light of 

 the crescent which is illuminated by the fun, and the light 



of the rest of the disc whicli is illuminated by the earth. 

 Sunlight exceeds full moonlight some 618,000 times, and it 

 exceeds full earthlight some .50,000 times. It is clear that 

 in the presence of a degree of brightness 50,000 times 

 greater, the brightness of the earth-lit part of the moon 

 must appear by contrast altogether insignificant. 



Then, again, we are apt to forget when we are looking 

 at a moonlit landscape, that the brightness of the scene 

 depends largely on the quantity of light, not on its in- 

 trinsic brilliancy. Let a small round hole be pierced at 

 the end of a blackened case ; let the observer be so covered 

 that he can see only through that hole ; and let the hole 

 be so placed as to correspond exactly in seeming magnitude 

 with the full moon : then, if the observer look through 

 that hole at a moonlit hill, he will at once see how utterly 

 insignificant moonlight is by comparison with sunlight. 

 And yet even then the comparison is made under circum- 

 stances altogether favourable to the moonlight. For under 

 the darkness in which the observer is enwrapped, the pupil 

 of the eye dilates, and a much larger proportion of light is 

 thus received than would be the case if no such change 

 took place. But let the experiment be renewed at the 

 time of new moon, and in such a way that only the light 

 from the earth-illumined part of the disc is in the field of 

 view ; then will the relative superiority of earthlight over 

 moonlight be clearly recognised. 



T do not speak at random. I have tried these experi- 

 ments. 



It would be easy to cite multitudes of similar instances. 

 But let us pass from the consideration of eye-estimations of 

 different degrees of light at different times, to consider how 

 far the eye is competent to compare different degrees of 

 lustre observable at the same time. 



It is very commonly stated, though not constantly, that 

 the planet Jupiter shows brighter at the edge all around 

 the disc, than in the middle. Certain observations of the 

 satellites serve to show that the reverse is really the case. 

 Noticing a passage in an excellent paper by Jlr. Browning 

 on the planet Jupiter, in which the former view was stated, 

 I expressed doubts on the point. " But," said Mr. Brown- 

 ing, "I do not )epeat anything I have heard or read on 

 this matter, I describe what I have seen." Knowing that 

 to prove himself mistaken would be quite as pleasant to 

 him as to prove himself in the right, I asked him to take 

 an early opportunity of testing the matter by the use of a 

 carefully-graduated darkening glass. He did so, and found 

 that though .lupiter (when in opposition, — near quadrature 

 there is a difference) look^- distinctly lirighter near the edge 

 than in the middle, he in distinctly brighter in the middle 

 than near the edge. The deception is due, no doubt, to 

 the contrast between the edi^e and the dark sky. 



I have very little doubt that the apparent brightness of 

 the moon's edge when she is full is to .win'' degree an effect 

 of contrast ; though it remains recognisable when the test 

 of the graduated glass is applied. 



And here another question suggests itself. What is the 

 effect of either intensifying or reducing in the same degree 

 all the lights of a surface variously illuminated ? Intensi- 

 fying is not easy ; since there is absolutely no device by 

 which any surface can be made to look brighter than it is* 

 Nevertheless, as we sometimes see surfaces less bright than 

 they really are, we may by removing the cause which had 

 diminished the light intensify the apparent brightness. 

 For instance, by travelling south, and getting to the 

 summit of a high mountain, an Englishman can gain 



• I am referring to celestial objects cliiofly. We may, of coarse, 

 illuminate a surface more or less brightly ; bnt extraneous lisht 

 causes perplexity in nearly all experiments on the illumination of 



terrestrial surfaces. 



