48o 



NATURE 



[December 8, 1921 



the process of manufacturing it is explained. This 

 gearing is of the straight-toothed type, and has been 

 developed by the Maag Gear Co., of Zurich. Means 

 have been found by which straight-tooth spur gears 

 can be employed successfully under conditions demand- 

 ing the highest possible peripheral speeds and loads 

 per unit width of tooth. The methods of production 

 ensure the requisite degree of accuracy, and at the 

 same time provide a tooth form which is considered 

 to give the most favourable conditions of sliding con- 

 tact during engagement, A novel grinding process 

 has also been devised whereby it is possible to generate 

 a correct profile on a case-hardened tooth surface by 

 means which are "independent of the wear which must 

 take place on the grinding disc. The compensating 

 arrangements in the grinding machine are extremely 

 interesting, and keep the grinding planes in their 

 correct position to within ±i/iooo mm. The gear 



has been applied to several ships and to a large 

 number of electrical plants and electric-locomotive 

 drives. 



"A Winter Guest/' the seasonable and striking 

 Christmas card published by the Royal Society for the 

 Protection of Birds (23 Queen Anne's Gate, S.W.), 

 represents the redwing, sometimes called the " Nor- 

 wegian Nightingale," though known in England only 

 as a winter visitor. The picture is reproduced in 

 colour from a painting by Mr. H. Gronvold, and 

 gives very happily the character and the appealing ex- 

 pression of the 'little traveller. It can be had, with 

 calendar, for 5^. by post. 



Erratum. — The Poets' Corner of Westminster Abbey 

 is in the south transept, and not the north, as stated 

 in the article on " Science in Westminster Abbey " 

 in last week's Nature, p. 437. 



Our Astronomical Column. 



The Effective Wave-Length of the Light of 

 Galactic Stars. — Prof. O. Bergstrand contributes a 

 paper on this subject to the centenary number of 

 Astron. Nach. The research is based on a series of 

 plates of the star-cloud in Cygnus obtained with a 

 15-cm. triple objective prism. The spectral types of 

 the brighter stars were taken from the Henry Draper 

 cataFogue, and the following table shows the correla- 

 tion between type and effective wave-length A. 



Type X Type A. Type X 



O 41 1/"/* F 421 K 429 



B 416 G 423 'M 439 



A 419 G5 425 



Only two M stars were available, both probably 

 giants. From the evidence of the G5 and K stars, 

 the author concludes that dwarfs have a smaller value 

 of A than giants of the same type. 



An analysis is then given of the values of A for the 

 fainter stars down to magnitude 13-4. It is pointed 

 out that the percentage of white stars is more than 60 

 for stars brighter than 11-4 m., and then drops sud- 

 denly to less than 50. He explains this by assuming 

 that the white stars have the highest absolute bright- 

 ness, and so are more distant than yellow or red stars 

 of the same apparent magnitude. Hence we reach the 

 limits of our local star s3-stem sooner in the white 

 stars than in the yellow ones, causing a drop in the 

 percentage for the fainter stars. Taking 00 as the 

 absolute magnitude of an average giant white star, 

 the radius of our local system is found to be 2500 

 parsecs. 



The Dynamical Equilibrium of the Stellar 

 System. — Prof. A. S. Eddington contributes an im- 

 portant paper to the centenary number of Astron. 

 Nach., in which he makes a further advance in the 

 solution of this problem ; his first approximation, in 

 which he took the shape of the system as spherical, 

 was published in Monthly Notices, R.A.S., vol. 75. He 

 passes on in the present paper to consider the much 

 more diflRcult problem of an oblate system, and 

 succeeds in finding one exact solution as follows : A 

 rotating system that has settled down to a steady 

 state may be presumed to be oblate. If we divide 

 it into two identically equal systems, and reverse the 

 direction of rotation of one of them, we shall have a 

 non-spherical system in equilibrium, with no rotation 

 as a whole, but with preferential motion in a trans- 

 verse direction. The solution is worked out in detail 

 for a homogeneous spheroid, which is shown to be 



NO. 2719, VOL. 108] 



strictly analogous to Maclaurin's hydrodynamical 

 spheroid. It is inferred that non-homogeneous solu- 

 tions exist. In conclusion, he points out that it is 

 quite likely that the stellar system has not yet attained 

 dynamical equilibrium, and may be collapsing some- 

 what rapidly towards a steadier condition. He has 

 hopes that the advance of knowledge of stellar masses 

 and velocities may enable the matter to be decided by 

 using the principle that in dynamical equilibrium the 

 total kinetic energy is half the exhaustion of potential 

 energy. 



Observations with the Cookson Floating Tele- 

 scope. — This instrument, which floats in a circular 

 tank containing mercury, was designed by the late 

 Mr. Bryan Cookson, and given after his death to the 

 Syndics of the Cambridge University Observatory. 

 It has been on loan at the Royal Observatory, Green- 

 wich, since 19 11, and a memoir just published dis- 

 cusses the observations of the first seven years. The 

 primary object was the redetermination of the aberra- 

 tion constant, while the variation of latitude is a 

 useful by-product. Pairs of stars are selected at about 

 the same distance from the zenith and on opposite 

 sides, and with right ascension difference of a few 

 minutes of time ; their magnitudes vary from 5 to 

 6-5. The trails of brighter or fainter stars are too 

 broad or too faint for accurate measurement. Three 

 solutions are given for the aberration constant and 

 for the corresponding value of the solar parallax as 

 follows : — 



Ab. const. Parallax 



20442 8815 General solution. 



20455 8-8io Discordant plates rejected. 



20460 8808 Correction for wind direction applied. 



The probable errors of aberration and parallax are 

 0-013" and o-oo6" respectively. The third solution 

 resulted from an examination of the "night errors," 

 which led to the conclusion that they depend largely 

 on the direction of the wind. This, combined with 

 the dissvmmetry of the ground-level to the north and 

 south of the instrument, is presumed to produce 

 inclination of the atmospheric strata and anomalous 

 refraction. It will be seen that the values for the 

 parallax do not differ too widely from Mr. Hinks's 

 value 8•8o6^ The period of observation is, however, 

 not considered quite long enough to eliminate acci- 

 dental error, and the observations are being con- 

 tinued for another seven years. 



