2 50 



NA TURE 



[January 13, 1898 



of the third type and is very red in colour, appears on 

 the print much fainter than S, while e Crucis, also a third 

 type spectrum star, which to the eye is so conspicuous 

 that it is usually inserted in rough diagrams of the con- 

 stellation, as for example on the Brazilian postage-stamps, 

 is on the negative so faint, that it would not be selected 

 as one of the principal stars. 



The trifid nebula in Sagittarius, taken with the same 

 telescope and exposure, illustrates still more completely 

 the inadequacy of this scale for star maps, but brings out 

 well the structure of such portions of the Milky Way as 

 fall on the plate. This particular region is remarkable 

 for the number of gaseous nebute that fall within the 

 area covered by the plate. One-tenth of the whole 

 number hitherto discovered in all parts of the sky are 

 here depicted ; but almost every night increases the 

 number of these objects in our catalogues. The ex- 

 amination of only two plates taken with the Bruce 

 telescope and objective prism showed that the 

 spectra of six known nebuke were gaseous. We pass over 



There seems to be no doubt about the place of the 

 nebulae so described, as a curved line of stars noticed by 

 Herschel is recognisable. But of the nebula itself there 

 is no trace. One seems to have another instance of 

 variability similar to Hind's nebula in Taurus. The 

 spectra of the stars contained in this portion of the sky 

 is also illustrated. By placing a prism in front of the 

 objective of the Bache telescope and exposing for 140 

 minutes, we have, exhibited on one plate, more than looo 

 spectra. Where the stars are densest the separate spectra 

 cannot be recognised in the print, but on the original 

 negative even these can be conveniently classified. No 

 less than fourteen stars within this limited area, out of a 

 total of sixty-three already catalogued, show a spectrum 

 which consists mainly of bright lines. 



The cluster known as w Centauri is admittedly the 

 finest in the sky. We reproduce, in Fig. 2, a reduced 

 form of the illustration of this superb object as it depicts 

 itself in the Boyden telescope after two hours' exposure. 

 In one particular this picture fails to do justice to the 



Fig. I. — Arequipa Station from the south-west. 



the photographs of the Magellanic clouds, reproduced 

 in the volume by contact printing, and reach the nebulous 

 region round r; Argus, or as Prof. Pickering prefers to 

 call it r) Carinte. This has been the subject of so many 

 illustrations that its main features are well known. It 

 serves here as an admirable object on which to test the 

 advantages of increased scale. Three different enlarge- 

 ments of the original negative are given, one taken with 

 the Boyden telescope. Prof Pickering apparently prefers 

 a scale of ten seconds to the millimetre at least for dis- 

 tricts where the stars are much crowded. This size 

 seems to him to possess the advantage of showing nearly 

 all the stars that can be seen upon the original negative, 

 without rendering the images inconveniently large. On 

 a careful comparison being made between the nebulae 

 recorded on the plate and those given in Dreyer's cata- 

 logue one very important difference was noted. N.G.C. 

 3199 was observed by Herschel on four nights, and on 

 each occasion described by him as very large and bright. 



NO. 1472, VOL. 57] 



original in the sky. The photographic images have 

 sensible diameters amounting to, perhaps, three seconds 

 of arc. This suggests the possibility that some stars are 

 hidden by those nearer to us, but, as Prof Pickering re- 

 marks, in the telescope each separate star can be 

 distinguished, and therefore it is not probable that 

 any star is actually occulted. The only two other 

 clusters in the sky, which though far behind, are to 

 some extent comparable with w Centauri in brilliancy, 

 are 47 Tucanae, and the well-known cluster in Hercules, 

 Messier 13. In each of these clusters Prof Bailey and 

 the staff have made attempts to count the number of 

 stars visible, and Prof Pickering has submitted the 

 results to a mathematical analysis. He finds that the 

 number of stars per square minute of arc increases in 

 arithmetical progression as the centre is approached, and 

 that the theoretical number derived from his formula 

 agrees fairly well with the actual count. In the case of 

 o) Centauri the number actually counted was 6336, while 



