September 9, 1920] 



NATURE 



59 



which show irregular curves of spectrum sensitiveness 

 have their minima raised, and in many cases smoothed 

 out entirely. It is preferred to add alcohol to the 

 ammonia bath, because without it the plates must be 

 used as soon as they are dry, although if it is omitted 

 the sensitiveness is increased to a still greater extent. 

 The communication is illustrated by numerous spec- 

 trum and other curves and diagrams, and some aerial 

 photographs in which a comparison can be made 

 between the effect of using the treated and the un- 

 ited plates. 



Prof. Luici Palazzo, director of the Meteorological 

 Office at Rome, has sent us a separate copy of his dis- 

 cussion in the Atvnali del R. Ufficio CentrcUe di Meteoro- 

 logia e Geodinamica (vol. xxxvii., part i) of comparisons 

 made of magnetic instruments at Terracina in 1901, 

 191 1, and 1913. On the first occasion the instru- 

 ments of the Meteorological Office at Rome were 

 compared with those used in the Polar expedition of 

 the Duke of the Abruzzi. On the other occasions the 

 comparison was between the Roman instruments and 

 others belonging to the Carnegie Institution of 

 Washington. Prof. Palazzo himself observed with the 

 Roman instruments. The American observers were 

 Mr. W. H. Sligh in 191 1 and Mr. W. F. Wallis in 

 19 13. Terracina was selected as being free from the 

 artificial disturbances which for some time have 

 rendered Rome unsuitable as a magnetic station. It 

 is interesting to note that all the dip-circles used were 

 of English construction. The magnetometers were 

 of different patterns. The comparisons with the 

 Washington instruments are of most general interest. 

 In the case of declination the mean difference between 

 the Rome and Washington instruments was only O'j' 

 in 1911 and 01' in 1913 — differences which are prob- 

 ably too small to rely on. In both years the 

 Washington instruments gave the smaller dip and the 

 larger value for the horizontal force. But the apparent 

 differences between them and the Roman instruments 

 were substantially different on the two occasions. 

 Consistency to o-i' in dip and to ly in horizontal force 

 is probably too much to expect from ordinary dip- 

 circles and unifilar magnetometers. 



The University Tutorial Press, Ltd., announces 

 that it has in preparation a new scries of text- 

 books for use in agricultural schools and colleges. 

 The first book of this series, which it is hoped will 

 be issued next month, is "The Chemistry of Crop 

 Production," by Prof. T. B. Wood, head of the 

 C.imbridge University School of Agriculture. This 

 will U- followed later by "Animal Food Production" 

 by ih., same author. A volume on "Chemistry for 

 Agricultural Students," by Mr, R. H. Adie, is also in 

 active preparation. 



W 1 l.;.irn from Mr. Wilson L. Fox that the state- 

 ni.ti^ 11 kI- in Nature for August 26 (p. 837) with 

 reUivuo- to the rainfall at Falmouth for August 18 

 is incorrect. The Daily Weather Report gives the 

 rainfall as 6 mm. between 7 a.m. and 6 p.m., and 

 not 2-31 in. between 8 a.m. and 7 p.m., as was 

 reported. 



NO. 2654, VOL. 106] 



Our Astronomical Column. 



The New Star is Cygnus.— Mr. Denning writes 

 that this nova declined in brightness very quickly 

 after its maximum on August 24, but at the end of 

 that month maintained its light for a few days, so that 

 its magnitude remained at about 42, and the nova was 

 a little brighter than the star 33 Cygni. The magni- 

 tude on August 24 was i-8. Since that date the 

 estimated magnitudes at Bristol have been : 



.August 



Radiation Pressure near the Sun. — Radiation 

 pressure has been freely invoked in recent years to 

 explain various celestial phenomena. Prof. Eddington, 

 who was one of the first to suggest that it may play 

 a very important part in the interior of the giant stars, 

 contributes a note to Mon. Not. R.A.S. for June, in 

 which hie examines the maximum effects that it could 

 produce outside the sun. Assuming a temperature of 

 10,000° C. at the surface of the photosphere (which 

 is purposely taken considerably in excess of the most 

 probable value), then the radiation just outside the 

 photosphere is equivalent to a wind-pressure of 30 dynes 

 per sq. cm., which could not support a greater mass 

 than I milligram against the solar gravitation. Thus 

 if a prominence 10,000 km. deep were upheld by 

 radiation, its mass could not exceed a milligram per 

 sq. cm. of its base, implying a density of 10- ", which 

 is difficult to accept, in view of the fact tliat the 

 prominences are visible in spectrograms as dark mark- 

 ings on the solar disc. Prof. Eddington estimates the 

 maximum density of the corona and the tails of comets 

 (on the radiation theon,') as 10-". These are less 

 difficult to accept, as it is known on other grounds 

 that their density is extremely small. 



DisTRiBirriON of Intensity in Solar and Steixar 

 Spectra. — Mr. Bertil Lindblad, of Upsala Observa- 

 tory, has made an important study of effective and 

 minimum wave-lengths in grating spectra of the sun 

 and stars (Upps. Univ. Arsskrift, 1920, Mat. och 

 Naturvetenskap 1). He claims that the method affords 

 an independent determination of absolute magnitude 

 without making use of the spectral lines, being thus 

 a check of the method of Adams and Joy. In par- 

 ticular, stars are giants the effective wave-length of 

 which exceeds 426 /x/n. The effective temperature of 

 tfie solar photosphere is found to diminish from 6200° 

 at the middle of the disc to 5430° at the limb. The 

 difference is ascribed to the lower depth from which 

 we receive light in the former case. Detailed 

 measures of the spectra of sixty stars of known 

 spectral type and luminosity are then given, and 

 curves drawn connecting minimum wave-length with 

 absolute magnitude for the various spectral types. 



A detailed examination of a special region in 

 Cepheus is then given, in which the data previously 

 acquired are utilised. It contains the extreme dwarf 

 KriJRor 60 ; two othor stars are found to be dwarfs, 

 B.D. 57° 2514 and 56° 2779. Their spectral parallaxes 

 arc given as 0-03' and 004*, and an annual P.M. of 

 0-5' is suspected in each case from comparison with 

 the B.D. places. 



On the other hand, there are several red giants in 

 the region, the distance of which is given a<i470o light, 

 years ; this is suggested as the distance of the galactic 

 star-cloud which orrupip<! the region. 



