October 12, 1905] 



NA TURE 



591 



Problems of Hereditj," reviews at length Mr. Archdall 

 Reid's recent book on the subject, and Miss Harriet 

 Munrop gives a picturesque account of a visit to Walpi to 

 study the snake dance. 



FiiOM a study ol the spectra of alloys of different 

 metals, photographed under varying conditions of electrical 

 *?xcitation, atmosphere, and the proportions of the com- 

 ponents, Mr. P. G. Nutting, of the Washington Bureau 

 of Standards, has arrived at some interesting conclusions 

 which confirm and supplement the results obtained by 

 I.ockyer and Roberts in 1873. Mr. Nutting's researches 

 are described in No. 2, vol. xxii., of the Astrophysical 

 Journal, and the results may be summarised as follows : — 

 The speitra of thr- component metals are independent of 

 one annihpr when the alloy is volatilised by either the arc 

 or the spark. The relative intensities of the component 

 spectra are unaffected by variations of the electrical con- 

 •ditions or by substituting hydrogen, o.xygen, mercury 

 vapour. &c., for air as the surrounding atmosphere. 

 Ceteris paribus, the spectrum of the component which has 

 the greater atomic weight will be the brighter, when in- 

 <luctance is used, either with the arc or with the spark. 

 Under certain conditions — which the author enumerates — 

 spectroscopic analysis of alloys to within an error of about 

 5 per cent, should be practicable. Mr. Nutting further 

 states that, in practice, the presence of impurities in the 

 electrodes is of little consequence, and that when alloys 

 are used as electrodes it is useless to attempt to intensify 

 the spectrum of either component by varying the con- 

 xlitions under which the arc or spark is produced. 



Tiui .American Academy of Arts and Sciences has pub- 

 lished a pamphlet giving a brief historical account of the 

 origin of the Rumford fund. This fund had its origin in 

 the gift by Count Rumford — who was born at Woburn, 

 Massachusetts — to the American Academy of .Arts and 

 Sciences of the sum of 5000 dollars, which was simul- 

 taneous with the gift of a like sum, 1000/., to the Roya! 

 Society. The purpose of the fund was the same in each 

 case, the award of a suitable premium for discoveries or 

 improvements in heat and light. The gift was accepted 

 by the academy, but for many years no award of the 

 premium was made, as no claimant appeared whose merit 

 was such in its opinion as to justify this. Meanwhile, 

 the fund had accumulated to the amount of 4000'., and 

 in view of the fact that there was no possibility of expend- 

 ing the income in the precise manner contemplated by 

 Count Rumford, application was made in 1831 to the 

 Supreme Court of the Commonwealth of Massachusetts for 

 relief, if such should be possible. The court issued a 

 decree which modified the possible disposition of the 

 income of the fund in such a manner as to increase its 

 usefulness while keeping entirely within the spirit of the 

 original gift. At the close of the last fiscal year of the 

 academy (1904-5) the Rumford fund amounted to 11,744/., 

 the income for that year having been 510/. A standing 

 committee of the academy known as the Rumford com- 

 mittee is charged with the supervision ttf the trust, and 

 considers all applications for the Rumford premium and 

 all applications made for grants in aid of research. The 

 Rumford committee was first constituted a standing com- 

 mittee in 1833. Its members were nominated annually 

 lj\' the president of the academy until 1863, since which 

 time they have been chosen in the same manner as the 

 •other officers. The Rumford fund of the Royal Society 

 has been devoted solely to the award of the premium 

 according to the original provisions of the trust. 



OUR ASTRONOMICAL COLUMN. 



Klrtuer Eclipse Results by French Observers. — In 

 No. 13 (September 25) of the Complcs rendus M. Salet 

 publishes the preliminary results obtained by his expedition 

 at Robertville (.Algeria) during the recent total solar 

 eclipse. 



M. Salet was in charge of the mission sent to this 

 station by the Bureau dcs Longitudes, the chief purpose 

 being to make researches regarding the polarisation of the 

 coronal light. 



The first point investigated was the existence of a 

 magnetic field in the neighbourhood of the sun, the presence 

 of such a field being evidenced by the deviation of the 

 plane of polarisation of the coronal light. The result 

 indicated that there is a very slight deviation, amounting 

 to about 2°. 5, which seems to show that in spite of its 

 great mass the sun has only a small magnetic field. 



.A photographic study of the distribution of the polarised 

 light of the corona showed the bands of polarisation de- 

 creasing regularly in intensity to about one and a half 

 diameters from the solar limb. The maximum of intensity 

 occurs at about 5' or 6' from the limb, and from this 

 distance the bands extend into the inner corona, diminish- 

 ing in intensity as they approach the edge of the moon. 

 A prominence extending across two bands shows no differ- 

 ence of intensity, thus proving the non-polarisation of these 

 features of the solar atmosphere. 



In order to test the atmospheric polarisation, two Savart 

 polariscopes were pointed go° from the sun, the one 

 towards the pole, the other towards the equator. At this 

 distance the quantity of polarised light during totality was 

 insufficient to observe, although at 30° or 40° from the 

 sun the bands remained visible throughout the period ol 

 totality. 



To determine the coronal, chromospheric, or atmospheric 

 nature of the corona spectrum lines, a " nicol " was 

 placed so that it covered half the slit of a spectroscope 

 and entirely extinguished the radially polarised light which 

 is reflected' by the corona. The resulting negative showed 

 that the continuous spectrum of the corona differed in 

 intensity on its two edges because of the suppression of 

 the reflected sunlight, but the light from a prominence 

 showed no diminution in intensity after passing through 

 the " nicol." 



The coronium line, which extends to about 4' from the 

 sun, is seen on both edges, as are the two calcium lines, 

 but the latter are stronger on the edge containing the 

 prominence. 



The ultra-violet region ol the spectrum, between X 338 

 and \ 305, was also photographed, and shows about fifteen 

 lines of which the nature and wave-lengths have yet to 

 be determined. 



CosMiCAL Evolution. — Some interesting results con- 

 cerning the processes of cosmical evolution are given in -a 

 mathematical discussion, by Mr. J. H. Jeans, of Cam- 

 bridge University, which appears in No. 2, vol. xxii., of 

 the Astrophysical Journal. 



The author first directs attention to the extremely small 

 densities usually obtained for such binary systems as that 

 of -Algol, and points out that these densities are incom- 

 patible with the assumption that such systems are com- 

 posed of incompressible homogeneous fluids ; but the 

 discussions concerning the mechanics of such systems are 

 primarily based on this assumption, and are, therefore, in 

 Mr. Jeans's opinion, deprived of any foundation of fact. 



Mr. Jeans then discusses the probable mode of evolution 

 of stellar and planetary systems, and arrives at the con- 

 clusion that " gravitational instability " plays the principal 

 part in the segregation of systems from the original nebula. 



He contends that Laplace's " rotation " theory of 

 cosmical evolution only takes into account a secondary 

 factor in the process, and, in support of the " gravitational 

 instability " theory, he shows that before rotation alone 

 could effect the birth of a satellite a nebulous mass of, say. 

 to"'" grams would have to contract until its linear 

 dimension was about 10-'''^ cm., i.e. until its density was 

 about 10"'. 



If the material of the original nebula could be con- 

 sidered as consisting of solid particles such as are assumed 



NO. 1876, VOL. 72] 



