88 



NATURE 



[April 4, 19 18 



that rainfall production is usually a complicated 

 process. 



The writer has not had enough high flying to 

 be able to make many observations of the upper 

 clouds, but they certainly present an interesting 

 field of investigation well within the possibilities 

 of aeroplanes. C. K. M. Douglas, 



PHOTOGRAPHIC DETERMINATIONS OF 

 STELLAR PARALLAX.^ 



THE determination of stellar parallaxes by 

 photography has shown a striking improve- 

 ment in recent years, and the results obtained 

 with the Yerkes refractor are of the highest order 

 of excellence. Two of the precautions observed 

 in this and similar series of measures are the 

 taking of all the plates at small hour-angles, so as 

 to minimise the effect of unequal atmospheric dis- 

 persion in the stars, and the reduction of the 

 magnitude of the parallax star to equality with 

 the comparison stars. This latter precaution is 

 necessary, since any inequality in the driving will 

 have a different effect on the images of objects 

 that differ much in brightness. The usual way of 

 effecting this is by rotating a screen in the form 

 of a sector of a circle in front of the brighter 

 image. By altering the angle of the sector, any 

 desired diminution of light may be obtained. 



This method was used for most of the paral- 

 laxes in the volume under notice, which, however, 

 mentions an alternative plan, due to Prof. 

 Kapteyn, that has been successfully tried at 

 Yerkes. It> consists in taking an out-of-focus 

 photograph of the required region, which on 

 development exhibits the stars as discs of equal 

 size but unequal density. This negative is then 

 used as a screen for the parallax plate. Since the 

 density of each disc is proportional to the photo- 

 graphic brightness of the star that formed it, it is 

 clear that the use of the screen will give nearly 

 equal magnitudes for the stars on the parallax 

 plate. 



The parallax work at Yerkes Observatory was 

 begun in 1903 by Dr. Schlesinger, who was ap- 

 pomted director of Allegheny Observatory In 1905 ; 

 it was continued by Messrs. Slocum, Mitchell, 

 Lee, Joy, and van Biesbroeck (of the Uccle Ob- 

 servatory, Belgium). Up to the end of 1915 131 

 parallaxes had been determined. The present 

 volume contains the details of the last eighty-five, 

 and a summary of the earlier results, which have 

 already been published. The parallax stars are 

 mamly bright ones, but nearly one-third of them 

 are famt stars with large proper motions. 



The parallax of Algol is given as 002'/ ; that of 

 61 Cygni, 027'/ ; of 70 Ophiuchi, 0-21'/ ; of € Lyra 

 (the double-double), 000'' (all four components 

 bemg measured); O.A. (N.) 17,415-6, 0-22//. 

 There are six of the eighty-five parallaxes between 

 o-i" and 0-2'', three above 0-2", and seventy-six 

 less than ci//. Four of the stars in the trapezium 



1 "Stellar Parallaxes derived from Photographs made with the 40-in. 

 Refractor. Publications of the Yerkes Ohservatory, vol. iv., part i. 



NO. 2527, VOL. lOl] 



of the Orion nebula were measured, as there is 

 little doubt that they are actually involved in the 

 nebula. The results are negative for all four 

 ( — 0-014", —0026'^ —0021'', —0023"), pre- 

 sumably indicating that the trapezium and nebula 

 are more remote than the comparison stars. The 

 possibility is recognised that the latter may them- 

 selves be involved in the nebulosity, and a further 

 investigation is suggested, using larger plates 

 that would include stars more distant from the 

 trapezium. 



The probable errors of these parallaxes are all 

 in the neighbourhood of o"oi". The error that is 

 reasonably possible is, of course, two or three 

 times as great. A good illustration of this fact 

 is afforded by the parallaxes found for the pair 

 of stars O.A. (S.) 14,318-20, R.A. ish. 5m., S. 

 decl. 16°. They are 5' apart, magnitudes g'6 and 

 9'2, spectral types G5 and G4, P.M.s 3'693" in 

 1957° ^nd 3*675 '^ in 195*6°, radial velocities 

 + 307 km. and + 295 km. These striking facts 

 leave no reasonable doubt that the two stars are 

 physically connected, and have sensibly equal 

 parallaxes. The present volume gives for the 

 parallaxes +0-025" +o-oo8" and +0-061" +0-012" 

 respectively. As Prof. H. N. Russell had previ- 

 ously obtained the values +0-014" +0-023",, 

 + 0045" +0-022", some astronomers have adopted 

 the view that one star is really some three times- 

 as distant as the other. But the close agreement 

 of their abnormally large proper motions renders 

 such a conclusion wildly improbable. In fact, the 

 weighted mean parallax is 0*040" from the Yerkes 

 plates, and 0*030" from those of Prof. Russell, a 

 quite satisfactory accordance. 



An appendix to the volume gives a detailed 

 description of the measuring machine in use for 

 these photographs. It was made by William 

 Gaertner and Co., Chicago. The screw is 18 mm. 

 in diameter, with 249 threads i mm. apart. The 

 nut is 50 mm. long, and the graduated head' 

 18 cm. in diameter, having 1000 graduations. 

 The errors of the screw are extremely small. 



A. C. D. Crommelin. 



THE SIKKIM HIMALAYA. 



NO section of the Himalaya is more fuU^ 

 known than Sikkim ; Kashmir even has not 

 been more assiduously investigated. The informa- 

 tion regarding Sikkim is important for two 

 reasons. This country, which extends, between 

 long. 88° and 89° E., from the Bengal plain to the 

 tableland of Tibet, is the only fully explored por- 

 tion of the eastern Himalaya. Our knowledge of 

 the more extensive territories of Nepal to the west, 

 and of Bhutan to the east, is relatively scanty. 



The pioneer explorer of this interesting land was 

 Sir Joseph Hooker seventy years ago. Since 

 1848-49 many others have studied its fauna and 

 flora, its geology and topography, its scenery and 

 people. Explorers, surveyors, collectors, members 

 of political missions, and expert mountaineers 

 have found in Hooker's "Himalayan Journals," 

 published in 1852, a pleasant companion and a: 



