514 



NATURE 



[March 31, 1898 



of these stars has been determined at "all the principal 

 observatories," to use Dr. Gill's phrase, who apparently 

 grew tired of enumerating all the institutions to which he 

 is indebted for the completeness of this section of the 

 work. The final value of the coordinates has been made 

 the subject of a discussion by Prof. Auwers, which may 

 well form a model for similar inquiries, and will be care- 

 fully studied by the professional astronomer engaged in 

 similar work. Into minute details which arise in this 

 section, as well as in the use and reduction of the helio- 

 meter measures themselves, it is impossible to enter here 

 with sufficient fullness to make the involved process at 

 all clear. The whole interest centres in the nicety with 

 which small residuals are treated. For a similar reason 

 it would be impertinent to offer any criticism which would 

 imply that we have given to the volume the same anxious 

 study and consideration which the combined authors 

 liave devoted to their subject. Dr. Gill has supported 

 liimself by the ablest authorities in meridional astronomy, 

 his own experience with the heliometer to which he has 

 devoted years of study in perfecting the mechanical 

 arrangements and details is profound, and we have no 

 doubt that we have here all that can be effected by 

 sagacity and experience in deriving the best results from 

 observations which are as perfect as we yet know how 

 to make them. The final outcome of the observations 

 which, in their main intention, were devoted to deriving 

 the value of the solar parallax may be thus presented. 



Heliometer f ^"^* discussed by Dr. Elkin.. 8'8i20 + 00090 



observations of ^•'^'^"^ " S""' ?ll • f^ois ± o-oo6i 



l^ Sappho ,, Dr. Gill ... 87981 + o'oii4 



Meridian j {."^ . " S""' a""^^'' f'F' Z°'^° 



observations ofl J''^^""^ " ^- ^"^^^"^^ \^^\ ± °"°5' 



l^bappho ,, Dr. Auwers 8 •626 +0"ii8 



The mean value from the heliometer measures is 

 5"'8o36 ± o"'oo46> while the meridian observations give 

 8""8o6 ± o""03o, but for reasons stated. Dr. Gill is in- 

 <:lined to adopt as a final value 8 "•802 ± o'oo5. Though 

 this result of itself would be a satisfactory outcome, the 

 accuracy of the observations permits some other astro- 

 nomical constants to be derived, either directly or through 

 their relations with other known constants. The Victoria 

 observations give with some confidence the mass of the 

 I 



moon = Q ,.^ J . For other constants, such as 



81702 ± 0094 ' 



the nutation, it is necessary to assume the luni solar 

 precession. The value adopted is 5o"'367 ± o""oo4, but 

 the source from which it is obtained is not very clearly 

 stated, neither is the epoch to which it refers. Ap- 

 parently it is taken from Newcomb's discussion in the 

 Astronomical Journal, No. 359 ; but in that paper we 

 have not been able to find this particular value, nor the 

 probable error with which it is accompanied. With this 

 value of the precession, however, and the exact amount 

 is immaterial for this purpose, the constant of nutation is 

 •9''''2o68, and pursuing the same line of inquiry the con- 



C-A 

 stant, —^ — (employing the ordinary notation), is 



o"oo32825. Adopting Clark's value for the equatorial 

 radius of the earth, the aberration constant is found to 

 be 2o"'467 ±^o"'oi2. Here as elsewhere the most prob- 

 able value of the solar parallax is assumed 8""8o2. 



The remaining volume, which contains the southern 

 " Durchmusterung," between the limits -18° to -37° 

 declination, is in its way quite as remarkable as the two 

 volumes which we have been considering. Herein we 

 have the first-fruits of the application of photography to 

 the determination of star positions on a large scale. The 

 old and the new methods are brought sharply into con- 

 trast. One would naturally like to institute a comparison 

 between the time necessary for the production of these 



NO. 1483, VOL. 57] 



zones, and that required for similar work, either at Bonn 

 or Cordova. But such a comparison is not easy, nor 

 probably would it be fair. The plates that were taken at 

 the Cape were measured at Groningen. Much time must 

 have been lost in correspondence and in settling the 

 details of a new method. Experience had to be acquired 

 in the most suitable methods of measuring with new and 

 untried apparatus. Prof. Kapteyn was necessarily oc- 

 cupied by his University duties, and could only devote his 

 leisure to the preparation of the catalogue — a leisure 

 which he gave unstintingly ; and Dr. Gill is to be con- 

 gratulated on the good fortune that supplied him with so 

 able and willing a coadjutor. 



The plates were taken with a rapid rectilinear Dall- 

 meyer lens of six inches aperture and fifty-four inches 

 focus. During the course of the work this lens was re- 

 polished, and a portion of the work duplicated with a 

 second lens ; but allowing for all interruptions, the work 

 that was begun on April 15, 1885, was finished in 

 December 1890, with the result that the whole sky was 

 photographed from the South Pole to - 19° declination. 

 The free area of each plate was five degrees square, 

 more than 600 being required to cover this portion of the 

 heavens once, without any duplication. At first, when 

 plates were rather slow in action, an hour's exposure was 

 given ; but this time was subsequently reduced to about 

 thirty minutes. It is instructive to notice that many 

 plates on a first examination had to be rejected, owing to 

 the fainter stars not having impressed themselves on the 

 film, on account of mist, dewing of the objective, or bad 

 definition. " The more thorough examination necessarily 

 made by Prof. Kapteyn in course of measurement, 

 brought to light a good many more plates which it 

 seemed desirable to re-photograph, so that some of the 

 areas have been photographed three, four, and even five 

 times." ■ Such a report will not be very satisfactory 

 reading for those engaged on the "Carte du Ciel." 



The measurement was effected in a manner that 

 necessitated very small corrections to the original read- 

 ings, in order to obtain the approximate star places, 

 referred to the equinox of 1875 > indeed. Prof. Kapteyn 

 says that the coordinates read from the instrument might 

 have been entered directly in the catalogue. We can 

 form a tolerably accurate notion of the time occupied in 

 measuring the plates, for it is stated that on good rich 

 plates two assistants could measure 300 to 400 stars in 

 an hour. Probably 200 would represent the average, and 

 since this portion of the catalogue contains 152,598 stars, 

 we have about 750 hours of actual measurement for one 

 complete examination. Such a rapid collection of results 

 needs no comment. The average distribution of the 

 stars throughout the whole area is possibly of greater 

 consequence than the actual number measured. Of 

 course, the number to a square degree varies very much 

 in different parts of the sky. In the sparsest parts, that 

 is, in Galactic Latitude about - 70°, this number falls 

 to 6*28, rather less than in Argelander ; but a comparison 

 of mean results with other zones gives the following 

 numbers. 



Cape Photographic Survey ... 25*43 stars to square degree. 

 Bonn N. Durchmusterung ... 15 19 ,, ,, 



Schonfeld i8*2i ,, ,, 



Thome, ... 56"i ,, ,, 



The arrangement of the stars in the catalogue, and 

 the degree of accuracy aimed at, is the same as in the 

 familiar Bonn work, namely one-tenth of a second of 

 time in R.A., and a tenth of a minute of decHnation. A 

 comparison between the places here given with those of 

 other catalogues shows that the probable error of a 

 photographic determination contrasts most favourably 

 with that derived from other processes. This i^ clearly 

 shown by the following table. -^ 



