August 21, 19 13] 



NATURE 



645 



investigation of the nature of the obscurity of the 

 air during the prevalence of the Harmattan wind at 

 Togo, West Africa, with reports (1) by Captain v. 

 Seefried, who conducted tin experiments with glass 

 filtering tubes packed with wadding, and (2) by the 

 Geological Central Section for German Protectorates. 

 The cloudiness has been supposed by Mime authorities 

 to be due to ashes from grass fires, and by 

 others to dust particles from desert regions. The geo- 

 logical report is to the effect that the principal cause 

 is essentially due to the remains of diatoms, and that 

 the presence of particles of ashes is only to be con- 

 sidered as a secondary and local phenomenon. It is 

 pointed out that this result agrees with opinions as to 

 the nature of the dust observed in the vicinity of 

 Cape Verde Islands, and it is suggested that some 

 botanist skilled in the examination of diatoms might 

 determine from their forms the place of origin of 

 the dust particles obtained in Togo. 



The Survey Department of Egypt has recently 

 published an important volume containing daily 

 readings (August, 1872-December, 191 1) of the Nile 

 gauge at Roda (Cairo), together with summaries and 

 diagrams. In a preface by Mr. J. I. Craig (director, 

 Meteorological Service) it is pointed out that projects 

 are in hand, or proposed, that will tax the supplies 

 of- the river to the utmost, and that it is more than 

 ever desirable that existing records should be made 

 accessible. With the exception of the nilometer at 

 Aswan, the gauge at Roda is the oldest in the 

 country; its readings have been recorded (with several 

 gaps) for about thirteen centuries, but it is only from 

 1872 that complete daily readings exist, and up to the 

 end of 1899 these were recorded according to the 

 Coptic calendar and in old measures, so that the 

 figures were of little value to European investigators. 

 The formidable work of conversion to Gregorian dates 

 and to modern measures, together with computations 

 for the whole period in question, has been carried 

 out with scrupulous care by Mohammad Effendi 

 Kasim (inspector, Meteorological Service), who has 

 also contributed the introductory text. This period of 

 daily records includes five cases of high floods (ex- 

 ceeding' 20 metres) and five cases of low floods (not 

 exceeding 18 metres), also years of abnormally early 

 and late floods, such as that of 1880 (forty-nine days 

 earlv) and that of 1894 (twenty-four days late). 



In a paper presented to the Astronomical and Astro- 

 physical Society of America in January, and reprinted 

 in abstract in the July Journal of the Washington 

 Academy of Sciences, Mr. W. J. Humphrys, of the 

 United States Weather Bureau, examines the question 

 whether the presence of volcanic dust in the upper 

 atmosphere is sufficient to account for the periods of 

 abnormal cold which have at times been experienced 

 simultaneously over the whole earth. He finds that 

 the meteorological records from 1750 onwards show 

 that cold periods were in every case preceded by vol- 

 canic eruptions, and that the greater the eruption 

 the longer the period of unusual cold. In the case 

 of the eruption at Krakatoa in 1883 the years 1884-5-6 

 were all i° or 2 F. below the normal temperature. 

 Measurements made at the U.S. Weather Bureau 

 NO. 2286, VOL. 91] 



since [883 show that dust in the upper atmosphere- 

 produced by an eruption may decrease bv 20 per cent. 

 tin- intensity of solar radiation received at the earth's 

 surface. 



OUR ASTRONOMICAL COLUMN. 

 A Star with Large Proper Motion. — In Astro- 

 nomische Nachrichten, No. 4674, Miss E. F. Bellamv 

 directs attention to the large proper motion of Hel- 

 singfors 914, which was found when assisting the 

 Vatican Observatory in tin- reductions of its portion 

 of tlu- Astrographic Catalogue (zones +65 to +55 ). 

 I he --tar is close to Hels. 913, and since the seconds 

 of R.A. in the Helsingfors Catalogue for 914 and 913 

 aie 448s. and 442s., whereas in BD they differ bv 

 5-.., and the Vatican residuals also differed by nearly 

 5s., it was at first thought that there was a mistake 

 in Helsingfors. Subsequent investigation showed that 

 Hels. 914 had a considerable proper motion, the 

 motions being + 0236s. and +036" annually. Miss 

 Bellamy points out that the centennial proper motion 

 in are of a gnat circle is +157-0", the largest proper 

 motion hitherto found in either the Oxford or Rome- 

 work, the next largest being 141-9". 



"Giant" and "Dwari " siaks. — In an address 

 given at the meeting of the Royal Astronomical Society 

 in June last, and reported in the current number of 

 The Observatory, Prof. H. N. Russell presented a 

 short account of the studies which have led him to 

 adopt a theory of stellar evolution through stages of 

 increasing and decreasing temperatures, as Ritter de- 

 duced from thermodynamical considerations and similar 

 to the idea of a "temperature curve" which spectro- 

 scopic work led Sir Norman Lockyer to suggest repre- 

 sented the course of development of a stellar body from 

 a sparse meteoritic swarm to a dense, feebly radiating", 

 and dving sun. Briefly, by considering absolute magni- 

 tude (a function of mass, density, and surface bright- 

 ness) and spectral type, Russell finds that there are 

 no faint white stars ; all the very faint stars are red, 

 and among the reddest stars, K and M classes, there 

 is a distinct separation into two groups, one much 

 brighter, the other much fainter, than the sun. To 

 explain the existence of the two kinds of stars, the 

 "giant" and "dwarf" stars first noticed by Dr. 

 Hertzsprung, recourse is made to the evidence afforded 

 by double stars and eclipsing variables ; from the 

 former differences due to mass are eliminated, and 

 from the latter the effects of variation of density are 

 deduced. It appears that the series of dwarf stars is 

 one of slowly increasing density from B to M, while 

 among the giant stars density decreases very rapidly 

 from B to M, and it is suggested that the giant stars 

 of class M represent a very early stage of evolution, 

 class B a stage near the middle, and the dwarf stars 

 later stages according to increasing faintness and 

 ruddiness. As a confirmation of the theory it is 

 pointed out that the actual densities of the stars of 

 class A and B are of the order of magnitude (one- 

 tenth that of water) at which Lord Kelvin predicted 

 temperature should be a maximum. 



The Stationary Radiation of Meteors. — The pre- 

 sent position of this puzzling question in the field of 

 meteoritic astronomy is discussed by Mr. W. F. Den- 

 ning in the August number of The Observatory. The 

 observational difficulties and possible accuracy of deter- 

 minations of radiants are indicated. The reality of 

 the phenomenon is insisted upon, and some of the 

 unsatisfying explanations which have been offered 

 bv Greg, Ranyard, Proctor, Herschel, and Turner, 

 among others, are briefly mentioned. 



