Ai'KiL 1 1, 1^95 



NA rURE 



559 



Photometer. In the first of these the principle of polar- 

 isation is still used, but the artificial star is discarded. 

 This is apparently an advantage, but it is a question if 

 it does not introduce an error as large as that which it 

 seeks to eliminate. An image of a star, as Polaris, is 

 used as the constant of comparison, and this image can 

 be reduced by polarisation till it equals that of the star of 

 which the magnitude is sought. A lack of resemblance 

 between the stars under consideration is removed, but 

 the removal is effected by the introduction of a second 

 object-glass with evidently different optical capacity, 

 requiring a fresh constant to be determined. Prof. 

 Pickering's photometer consists practically of two tele- 

 scopes, placed at right angles to the meridian, and over 

 each of the object-glasses is placed a right-angled prism. 

 By means of the northern prism the image of Polaris 

 is reflected, and by suitable adjustment can be made to 

 occupy any convenient position in the field of view, 

 while the prism on the other object-glass can be set to 

 any declination so as to bring the image of any other 

 star, when on or near the meridian, into ju.xtaposition 

 with that of Polaris. Ingenious arrangements are intro- 

 duced to ensure the coincidence of the pencils forming 

 the images to be compared, and a control over the 

 accuracy and efficiency of the whole is secured by con- 

 trasting the brilliancy of Polaris with itself — that is, by 

 comparing the images formed by either object-glass. 

 This is effected in all cases by rotating a Xicol prism in 

 the eye-piece of the telescope through a measurable 

 angle, and thus equalising the lights of the stars by 

 means of varying the planes of polarisation. How 

 effective this instrument has proved itself in the hands 

 of Prof. Pickering, we shall presently see. 



But either of these forms of photometer is necessarily 

 a special production, and therefore the object-glasses are 

 small and the light-gathering powers limited. In Prof. 

 Pickering's first photometer, the aperture was only 4 cm., 

 with a magnifying power of only fifteen diameters. Prof. 

 Pritchard, considering this limitation a defect, directed 

 his attention to the construction of a photometer which 

 should be readily available on any instrument, and be 

 applicable to stars of very varying degrees of brightness. 

 For this purpose he had recourse to the principle of e.<- 

 tinction of the light of a star, by means of a wedge of 

 neutral-tinted glass, which could be moved over the image 

 of a star till its rays were lost by the gradual increasing 

 thickness of the medium through which they had to 

 penetrate. This principle had been used by the late 

 Mr. Dawes, and also by Capt. Abney, but the long-con- 

 tinued use of such an apparatus by the late Savilian 

 Professor is likely to connect his name with this form of 

 photometer. The main defects in its construction arise 

 from the difficulty of obtaining an absorbing medium 

 equally operative throughout the entire length of the 

 spectrum, and also that of determining with certainty 

 the coefficient of absorption — in other words, how 

 much light is lost by the difference of thickness cor- 

 responding to one inch in length of the wedge. Recent 

 and more exact methods than those employed by Prof. 

 Pritchard seem to show that the constant used in his 

 work is in error, and that a correction to his magnitudes 

 so obtained is necessary. But it is a peculiarity of the 

 form of construction and method of observation adopted 

 that such a correction can be very easily applied. 



These forms of photometers, the/Collnor, the Pickering, 

 and the Wedge, are the three instruments which have 

 been most generally in use, and with which the modern 

 work has been accomplished. The rapidity and the 

 progress of this class of observation can easily be shown 

 by a few statistics. Previous to their introduction, exact 

 photometry was limited practically to two catalogues. 

 Exact photometry is, of course, a relative term ; it is 

 meant to include observations other than eye estimations, 

 and therefore Herschel and Seidel, the one with 69 stars, 



NO. 1328, VOL. 51] 



the other with 208, are the only two observers to whom 

 it is necessary to refer. Since the introduction of the 

 more rapid methods, and possibly from a better appre- 

 ciation of the importance of the inquiry, we have had 

 many extensive catalogues. Leaving out Zollner him- 

 self, who did not attempt to condense his observations 

 into catalogue form, we have — 



Peirce's Harvard CataIo5;ue of 494 stars. 

 Wolff's First Honn Catalogue of 475 ,, 

 Wolfl's Second Catalogue of 923 ,, 



Potsdam Photometric Catalogue of 3522 ,, 



All these catalogues have been made by means of a 

 Zollner photometer, but the list in no way e.shausts the 

 photometric work that has been done by this instrument. 

 For instance, Lindemann, at Pulkova, has carried out a 

 long series of observations with the view of determining 

 the scale that has been unintentionally adopted in the 

 record of eye estimations in various catalogues. Ceraski 

 and others have been at work on variable stars, while 

 interesting inquiries into the extinction of light by the 

 atmosphere and other physical investigations have been 

 made by its aid. A debt of gratitude, therefore, of no 

 common kind is due to the ingenuity of Dr. Zollner. Con- 

 fining our attention, however, solely to the compilation 

 of catalogues, we have with the Pickering meridian pho- 

 tometer a collection of the relative magnitudes of 4260 

 stars, followed by a photometric revision of the Durch- 

 musterung of Argelander, in which are given the magni- 

 tudes of some 17,000 stars. This leaves out of the 

 summary a quantity of miscellaneous work on the 

 Pleiades, on the .•\steroids, on double stars, standard 

 stars, &c. In fact. Prof. Pickering has placed it on 

 record, that the number of measures made with the 

 Nicol prism was up to 1890 slightly under half a million. 

 Finally we have the more modest catalogue of Prof. 

 Pritchard, embracing 2647 stars, and, to complete the 

 record with this particular instrument, we must add a 

 small item of some 45,000 extinctions made at Harvard 

 under Prof. Pickering's direction. Of course, many stars 

 are common to all the catalogues, but the record shows 

 I that in the last few years instrumental photometry has 

 been applied to something like 30,000 stars. It is not 

 easy to form an adequate conception of so much activity. 

 But if the numbers have increased in such welcome 

 proportions, it may be asked is there an equally gratify- 

 ing advance in the accuracy of the observations .' This 

 question is not so easily answered. Doubtless there is a 

 much greater accord among the observations found in 

 the same catalogue, and made by the same observer, 

 employing the same instrumental means. But when 

 these catalogues are compared with one another, 

 large and provoking differences are sometimes en- 

 countered, and not a small portion of time has been given 

 by various astronomers to the investigation of these dif- 

 ferences, and the attempt to systematise the various re- 

 corded values of lustre. But when all has been done, there 

 still remain individual differences which baflleexplanation. 

 They seem to point either to irregular variations of bril- 

 liancy in the stars themselves, or to baftling meteoro- 

 logical influences, which it is impossible entirely to elimi- 

 nate. The suggestion has been made by others, and it is in- 

 tended here to give it the fullest support, that a far 

 larger number of stars exhibit variations of lustre than 

 are included in our variable star catalogues. It must be 

 remembered that these catalogues have been formed, 

 and the variations detected, by the eye alone— that is to 

 say, without the advantages of a photometer. Conse- 

 quently it is only the larger variations that have attracted 

 attention. It is not easy to establish the fact of an 

 alteration in brilliancy, if it be small, either with or with- 

 out a photometer ; but it seems not unlikely that as star 

 magnitudes gain in trustworthiness, a larger addition 

 will be made to the stars recognised as variable. To 



