July 21, 1892] 



NATURE 



275 



the ratio to express the relative brilliancy between two 

 consecutive magnitudes. A committee was appointed to 

 carry this plan into execution, but the report which this 

 Committee has issued is unfavourable to the adoption of 

 the method. The signatures of the Astronomer Royal, 

 Prof. Pritchard, and the brothers Henry, are attached to 

 this report ; but M. Vogel, the remaining member of the 

 Committee, has not found the reasons assigned by his 

 colleagues sufficient to warrant the rejection of the 

 scheme, and consequently his name does not appear. 

 The President of the Permanent Committee thus sums 

 up the case against the proposal. Light in traversing a 

 metallic screen of bright threads and very narrow mesh, 

 seems to experience, besides the ordinary effects of 

 diffraction, certain modifications, whose cause is not yet 

 explained, and which the Congress could not foresee when 

 they framed the recommendation. This peculiar be- 

 haviour of the light demands further study, and renders 

 the application of this means very difficult, if not useless, 

 for the purpose for which it was proposed, since the 

 discrepancies of the results obtained are greater than the 

 error that an experienced astronomer would make in 

 estimating stars of the eleventh magnitude. 



The experiments on which this conclusion is founded 

 are set out in considerable detail, and a careful study of 

 these experiments ought to convince an unprejudiced 

 critic that the committee was justified in advising the re- 

 jection of the screens as an adequate and efficient means 

 of deciding upon stars of the eleventh magnitude. It 

 should be stated that the gauze screens, identical in char- 

 acter, were furnished by Prof. Vogel, and though there is 

 no mention of the experiments or processes which induced 

 the Potsdam astronomers to select a screen of this par- 

 ticular obstructive power, it is to be presumed that in his 

 photographic telescope they stopped the amount of light 

 proposed by the Congress. It is not the least curious 

 feature in the discussion (controversy would be far too 

 strong a word to describe the courteous paragraphs in 

 which the various astronomers set forth their reasons 

 for dissent from the able physicist), that Prof. Vogel takes 

 no part in it nor vouchsafes any information as to the 

 principles by which he was guided in the selection, but 

 leaves the onus of rejection entirely to his colleagues, who 

 are thus placed at a disadvantage. 



Prof. Pritchard, whose photometric researches permit 

 him to speak with authority, has stated concisely the 

 result of his experience. He found that on the ordinary 

 astronomical telescope, achromatised presumably for D, 

 the amount of light obstructed was equivalent to i'\ mag., 

 and on the photographic telescope, with a minimum focal 

 length for G, the amount of light lost was not less than 

 2"8 mag. The Astronomer-Royal reports that the action 

 of the screen on the Greenwich telescope is to stop 2'5 

 mag. This result was deduced by comparing the seventh 

 and ninth magnitude stars of Argelander. Some further 

 comparisons of the obstructed and unobstructed light of 

 stars of the ninth and eleventh magnitude photometrically 

 examined by Prof. Pritchard with the wedge photometer 

 confirmed this result, and further proved that the scale of 

 Pritchard and Argelander was in very satisfactory and 

 close agreement. It will be necessary to return to this 

 point. M. Henry at Paris offers results in close accord- 

 ance with those of the two English astronomers just 

 quoted. He finds that the screen proposed by M. Vogel 

 as effective in his instrument stops between 2"5 and 27 

 mag. on the Paris telescope, and this effect is still further 

 confirmed by some observations by M. Trepied, while M. 

 Rayet at Bordeaux finds 27 mag. represents the effective 

 action of the screen. Very different is the experience of 

 M. Donner, of Helsingfors. His method of estimating 

 the loss of light is different from that employed in the 

 other cases, and is perhaps not without objection, but 

 the result which he derives from his observations is that 



NO. 1 186, VOL. 46] 



the light of a star in passing through the screen loses 

 only 1 "6 mag. 



It is now necessary to describe very briefly the methods 

 employed in the various observatories which have led to 

 these discordant results, the more so as one eminent 

 authority. Dr. Duner, of Lund, who apparently holds a 

 brief for Prof. Vogel, has taken exception to the results 

 deduced. Leaving on one side the experiments con- 

 ducted by MM. Henry and Trepied on artificial stars, 

 and against which Dr. Dundr urges no objection further 

 than that they are founded on artificial stars, we find that 

 one principle pervades the examination conducted at 

 Greenwich, Paris, Bordeaux, and Algiers. The several 

 astronomers have determined what length of time is 

 necessary to produce a blackened star disc of the same 

 diameter from the same star with and without the screen. 

 In this way it has been found necessary to expose for ten 

 or eleven times as long with the screen before the object 

 glass as without, and from this fact it has been inferred 

 that the loss of light occasioned by the screen amounts to 

 2'5 or 2-6 mag. It is urged that if only two magnitudes 

 were lost by obstruction, the necessary exposure would 

 have been (2-512)" = 6-3, that required by the unobstructed 

 object glass. Dr. Dundr remarks on this that those who 

 have condemned the employment of the screens on these 

 grounds have argued in a vicious circle, and to be logically 

 correct it would be necessary to show that the intensity 

 varies as the time of exposure or 



Against the accuracy of this law Dr. Dun^r urges that 

 reports of the observers themselves show three distinct 

 proofs. In the first place (i) Dr. Donner states that only 

 o'58 mag. is gained by successively multiplying the length 

 of exposure by 25 ; (2) that the Astronomer Royal proves 

 that a gain of 17 or 1-85 mag. is secured by multiplying 

 the length of exposure by 6-25 ; and (3) that MM. Henry 

 have found that to obtain similar discs from stars of the 

 9'3 and 11 '3 mag. the exposure has to be increased from 

 28sec. to 2403ec. (i : 8-6). These three experiments give 

 instead of 2-512 respectively, 



3-28, 2-69, 2-93. 

 results apparently incompatible with the formula 



iV= const. 

 MM. Trt^pied and Henry reply at length and effectively 

 to these strictures. They do not regard 2-69 and 2-93 as 

 differing so greatly from 2-512 but that the discrepancy 

 may be fully explained by inaccuracy and paucity of 

 observations. The Helsingfors result (3-28) they refuse 

 to accept as unequivocal evidence in the face of estab- 

 lished facts. The method of Dr. Donner consisted in 

 comparing photographs of the Pleiades, taken with and 

 without the screen, with the map of M. Wolf, and mark- 

 ing the number and magnitude of the star.s which have 

 black or grey images. This method, as already hinted, 

 does not seem to be entirely free from objection. Ad- 

 mitting that the comparison of the images was made, as 

 we are sure it was, with all the care possible, there is still 

 room for the varying exercise of individual judgment as 

 to what constitutes a black and what a grey image, and 

 the final result is likely to be less exact than a process 

 based upon rigorous measurement. 



The method employed by Prof. Pritchard is, perhaps, 

 as free as any from objection or misinterpretation. He 

 exposed the plate for equal times with and without the 

 screen, and then measured the diameters of the resulting 

 star discs. If two discs, produced, one with, and one 

 without the screen, were found equal in diameter, then 

 the effect of the screen is equivalent in photographic 

 action to the original difference of magnitude between 

 the two stars. This difference of magnitude was deter- 

 mined by the wedge photometer, and the only exception 



