November, 1911. 



KNOWLEDGE. 



427 



when adapted to twilight vision. The eye, in fact, aeeording 

 to this view, iiiMst be adapted to twilight, or must have the 

 so-called " Dark-adaptation," before the phenomenon can be 

 produced. The retinal apparatus of the eye undergoes a 

 certain change when dark -adaptation occurs, the rods of the 

 retina, which are distinguished from the cones by the posses- 

 sion of a substance peculiarly sensitive to the action of liglit 

 known as " visual purple," being more particularly in\ohed 

 in the process. 



So nmch, then, for the physiulo^ical explanation of those 

 curious changes in the rclatise brightness of colours at dusk, 

 known as the " Purkinje Phenomenon," and this is, in essence, 

 the explanation given by Dr. Charles S. Myers in his 

 recent science manual ",\n Introduction to Experimental 

 Psychology""; but it appears to nie that there are other 

 factors of a purely physical character — factors which I have 

 so far never seen adduced — which also conspire to bring .about 

 the remarkable effects observed. 



Every photographer must be familiar with the extraordinary 

 change of aspect which coloured objects suffer when seen 

 under the red rays of the dark room. Reds appear very 

 bri,'lit, and blues correspondingly dark, for he is regarding 

 them by what is practically monochromatic light of a red hue, 

 and only those objects whose surfaces are capable of reflecting 

 the red light waves will show brightly: all others must appear 

 dark in comparison. Now. when the Sun has set, we have the 

 reverse phenomenon, for our source of illumination is the blue 

 sky only. The Sun, which sends us rays of all wa\e-lengths, 

 being below the horizon, surrounding objects can only 

 be rendered \isible by the sunlight reflected from the 

 minute particles of the atmosphere, and these, as we 

 know, send us predominantly light of short wave-length, 

 so that we here again have a quasi-monochromatic effect. 

 This bluish light, therefore, will be strongly reflected by all 

 objects of a blue tint, rendering them brightly conspicuous, 

 while, conversely, all red objects will appear correspondingly 

 dark — we shall have, in point of f.ict, a displ.iy of the 

 Pmkinje Phenomenon. 



While in no way venturing to impugn the validity of the 

 physiologic.d interpretation of the phenomenon, as first 

 described, it seems to nie that we possess in the purely 

 physical aspect of the question an equally powerful, if 

 generally neglected, factor contributing towards the identical 

 result. ' W. .\LFK1£D PARK, 



n.\ki:d-i-:vi-: comp.ts. 



T(i the Editors of "Knowledge." 



Sirs. — The fact that four comets have been visible to the 

 naked eye within two months must furnish a fact of exceptional 

 rarity. I saw Kiess's comet on .August 3rd without telescopic 

 aid. Hrooks's on .August 15th. and Quenissett's on September 

 2Sth. .And during the past few mornings Beljawsky's comet 

 has been very generally seen as a conspicuous object in Leo. 



Brisiol October Sth. '^^'- F- Dl^NNINHi. 



ASTRONOMICAL nU L.RIES. 

 To the Editors of " Knowledge." 



SiKS, — May I be allowed to thank Dr. Crommelin and the 

 Rev. M. Davidson for their lucid replies to my (juestions in 

 your issue of August, and to add some further remarks ? 



On the first two (juestions I have nothing further to say, 

 except to express surprise that the Nautical Almanac should 

 publish the time of greatest brilliancy of Venus to the nearest 

 hour when such accuracy is of no significance. In this 

 connection I should like to refer to the ease with which the 

 planet was seen at Hampstead at inferior conjunction on 

 September 15th, when within nine degrees of the Sun himself. 



With reference to my third question, dealing with the Moon's 

 secular acceleration, I had no intention of submitting an 

 astronomical catch but was in a real difficulty, though I found 

 the fallacy in my own argument shortly afterwards. My 

 solution, which agrees in its results with that of .Mr. Davidson, 

 is as follows ; 



* Ciuiitjiidge Uni\ 



Taking D. the disturbing acceleration (using the same 

 notation as Mr. Davidson), as proportional to the reciprocal 

 of the third power of the earth's radius \ector. 



Then the a\-erage value of D throughout a complete 

 revolution of the earth is 



1 



1- - f 'it , ,. . 



D = K / - . . where K i.-, a consi.u 



T J „ R' 



it. 



Hut (/ '^ -"- - d 0, wIk.tl- <I> is the ciiiiipkiiiciit of thr 

 K 



rt (1 — c I 



From this it appears that, as e decreases and D' conse- 

 quently decreases, the Moon will be accelerated. Hut yet I 

 am not satisfied. The problem is to show that the ai'era^e 

 angular velocity of the Moon in her orbit increases as the 

 eccentricity of the Earth's orbit decreases. The above shows 

 that the average perturbing acceleration decreases with the 

 eccentricity, and we know that the angular velocity increases 

 with decrease of perturbing acceleration ; but I submit that 

 these are not sufficient to prove the proposition. It seems 

 possible that there may lurk here a fallacy similar to that with 

 which I began this correspondence. 



Perhaps my meaning will be clearer if I put the matter into 

 symbolic form. The angular velocit>'. w, being a function of 

 D, we have w = '/' (D) say. 



„- , ,, i d( (D)) . 



W e know that ; is negative. 



d D 



(1) 



Wc ha\e found that d ( / D dt ) is positive (21 



de 



But does it necessarily follow that 



■■A IDI dt I is negative? (.i) 



de 

 We still require to know the relationship a ~ >(■ IDI. 



1 do not. of course, doubt the well-recognised truth of (31, 

 but 1 venture to think that it does not follow from (1) and (2). 



C. O. HARTRU.M. 

 eisil)* l'i'es>, 191 1. 



