656 



NATURE 



[May 20, 1922 



now has 400 members, 250 of whom are fellows. 

 The income from subscriptions was a little over 500/. 

 and the office salaries 500/. During the year the 

 Board formulated a scheme for the pubHcation of a 

 Journal of Scientific Instruments, and the 'Institute 

 has received a grant of 250Z. from the Department of 

 Scientific and Industrial Research to enable it to 

 produce the first number. The Institute has provided 

 50/. for the expenses of distribution of the 10,000 

 copies which have been printed, and in a short time 

 the number will be in the hands of those likely to be 

 interested in such a Journal. The price will be 

 30S. per annum if sufficient support is provided 

 by instrument makers, research associations, scientific 

 societies, and the public to justify the regular issue 

 of such a periodical. 



The Board of Education has issued a memorandum 

 on the effect of the Summer Time Act on the health 

 of school children. All Tocal Education Authorities 

 in England and Wales were circularised in May 1921, 

 and only 16 authorities, representing about 127,000 

 children, failed to reply. Of the 299 authorities 

 from which replies were obtained, 183 authorities, 

 representing 3,227,842 children, are definitely in 

 favour of the Act ; 89 authorities, representing 

 1,600,429 children, consider the Act detrimental, 

 while 27 authorities, representing 232,402 children, 



have not formed a definite opinion. It should be 

 added that nearly all authorities who hold that the 

 Act is prejudicial to children, state that this is due, 

 not to any defect inherent in the Act itself, but to 

 the fact that parents do not use it rightly. If 

 parental control were properly exercised nearly 

 every authority in the country would approve the 

 Act. The actual (though not the necessary) result 

 of the Act appears to be that large numbers of children 

 lose a valuable hour of sleep, because they go to bed 

 at dusk as before, but have to get up an hour earlier, 

 as the working part of the family rises by the clock. 

 The Board is therefore issuing a circular to the 

 parents of school children, pointing out how important 

 adequate sleep is for the growing child, stating the 

 amount of sleep necessary at the different ages, and 

 supporting the arguments by some common-sense 

 appeals to the parents. 



It is announced that the British Association for 

 the Advancement of Science will publish, early next 

 month, " The British Association : A Retrospect, 

 1831-1921," by the Secretary of the Association, 

 Mr. O. J. R. Howarth, which will present a summary 

 review of the activities of the association in every 

 department of science since its foundation. The 

 production of this volume has been rendered possible 

 through the generosity of Sir Charles Parsons, ex- 

 president, at whose suggestion it was undertaken. 



Our Astronomical Column. 



A New Variable in Cygnus. — The star B.D. 

 + 34° 4217 (position for 1900: R.A. 2oh. 54m. 12s. 

 Decl. + 34° 47'-4) has been discovered by Mr. 

 Stanley Williams to be a short -period variable 

 {Monthly Notices, R.A.S., vol. 82, p. 300). He com- 

 menced visual observations on this star with a 6J-in. 

 reflector in October last, and has deduced a light 

 curve which at first appeared to be of an Algol 

 type, but later was found closely to resemble that of 

 ^ Lyrae. The period is about i5h. 9m. and the 

 range of magnitude 10-42 to 9-93, the magnitude at 

 secondary minimum being 10-15. 



The Spectrum of the Corona in 191 8. — The 

 expedition sent from the Lowell Observatory to 

 observe the eclipse of June 8, 191 8, was stationed 

 near Syracuse, Kansas, and obtained some interesting 

 results in spite of rather unfavourable weather. The 

 equipment consisted of two single -prism and two 

 three-prism spectrographs (one being a slitless instru- 

 ment). The distribution of coronium was found to 

 be very different from that of hydrogen and helium. 

 The line at X5303-0 showed that it extended to a 

 distance of about one solar diameter above the sun's 

 surface, and the condensations in the green coronium 

 ring as shown by the objective-prism plates indicated 

 a distribution which was not in any way related to 

 individual prominences. There appeared to be a 

 general correspondence in the distribution of coronium 

 and the main features of the corona, since it was faint 

 or absent in the regions occupied by the polar 

 " streamers " and abundant beneath the main 

 extensions of the corona. The arches over pro- 

 minences were unusually well developed, but the 

 presence or absence of coronium in them could not be 



NO. 2742, VOL. 109] 



verified. The principal results are described by 

 Slipher in the Astrophysical Journal, 55, p. 73. 



Determination of Luminosities by Spectro- 

 photometry. — Two new methods for determining 

 stellar absolute magnitudes are described by Lind- 

 blad in the Astrophysical Journal, 55, p. 85. The 

 first of these is applicable to stars of higher spectral 

 type than those for which Adams's spectroscopic 

 method is available, and is based on the variations, 

 with absolute magnitude, of the energy distribution 

 of the spectrum between He and Hf . The two regions 

 X\3889-3907 and AX3907-3935 are compared, and 

 faint stars found to show a considerable relative 

 decrease in intensity of the former region when 

 compared with bright stars. This is probably due 

 to the widening of H^ and of some arc lines of iron 

 and silicon. The actual measurements are made by 

 taking a series of exposures of each star on the same 

 plate with decreasing exposure-times. In the series 

 of images thus obtained two are selected such that 

 the intensity of XX3889-3907 in one is equal to that 

 of XX3907-3935 in the other ; and the ratio, E, of 

 the two exposure-times is plotted against the absolute 

 magnitude, M, of the star. Curves are given showing 

 the relation between log E and M for different types, 

 and it is claimed that for types B^-Aj absolute 

 magnitudes may be found (with the dispersion used) 

 with a probable error of only +0-4 mag. 



The second method is similar to the first, but is 

 only applicable to stars (types G-M) in which there 

 is appreciable cyanogen absorption. The relative 

 density on either side of X3889, and between the 

 regions XX4144-4184 and XX4227-4272 are found and 

 compared as above with the absolute magnitudes 

 of the stars. 



