June 24, 1920] 



NATURE 



529 



dry- and wet-bulb temperature, relative humidity, 

 cloudiness, sea-surface temperature, strength of cur- 

 rent, and rainfall, with notes of any unusual pheno- 

 mena, including the appearance of albatrosses and 

 schools of dolphins or flying-fish. The charts included 

 with the publication show the salinity and tempera- 

 ture in depth sections, the one for Mogador giving 

 salinity, temperature, and density separately. The 

 salinity seems to decrease southwards, and also 

 generally with increasing depth. 



The March number of Terrestrial Magnetism and 

 Atmospheric Electricity contains a summary, by Mr. 

 J. P. Ault, of the results of the magnetic survey of the 

 Atlantic made by the Carnegie during her voyage 

 from Washington to Dakar, West Africa, and Buenos 

 Aires during the autumn and winter. While the 

 values found for the magnetic dip differ often by two 

 or three degrees from those given in the last Admir- 

 alty Charts 3598, 3603, and 3775, the values of the 

 observed deviation of the compass to the west differ 

 by more than a few tenths of a degree from the 

 charted values in certain limited regions only. Thus 

 in the region between the Gold Coast and the Island 

 of Ascension the Admiralty Chart gives the deviation 

 to the west about one degree tJo large, and between 

 Trinidad and Buenos Aires there is a considerable 

 area in which the deviation is given too small by 

 the same amount. 



Owing to the decrease in research at Harvard dur-^ 

 ing the war, vol.xiii. of Contributions from the Jeffer- 

 son Physical Laboratory covers the three years 

 1916-7-8, and at least a third of the volume is devoted 

 to Dr. P. W. Bridgman's work on the effects of pres- 

 sure on the electrical resistance and thermo-electric 

 properties of niore than twenty metals. The pressures 

 used reach 12,000 kilograms per sq. cm., and the tem- 

 perature ranges between 0° C. and 100° C. With the 

 exception of wires of bismuth and antimony, the 

 resistances of metidlic wires subjected to hydrostatic 

 pressure decrease with the pressure, following a linear 

 law approximately, and at 10,000 kilograms per sq. cm. 

 have values about 99 per cent, of their values at 

 atmospheric pressure in the case of cobalt and tung- 

 sten down to about 90 per cent, in the case of lead, 

 tin, and cadmium. The temperature-coefficient of 

 resistance remains almost unchanged. The effect of 

 pressure on the thermo-electric properties is much 

 more variable. The normal effect is to increase the 

 thermo-electric power of the metal, but in three out of 

 the twenty metals tried this is not the case. In most 

 cases both the Peltier and the Kelvin effects are in- 

 creased, but there are many exceptions. The author 

 considers that the electron theory is quite incapable of 

 explaining these results. 



We have received from Messrs. C. Baker, High 

 Holborn, W.C, their classified list (No. 69) of 

 second-hand scientific instruments. The list includes 

 microscopes and accessories, telescopes and field- 

 glasses, spectroscopes, surveying, astronomical, pro- 

 jection and physical apparatus, and contains particu- 

 lars of more than 2000 pieces of apparatus. In these 

 days of high prices intending purchasers would be 

 well advised in the first place to consult Messrs. 

 Baker's catalogue. 



NO. 2643, VOL. 105] 



Our Astronomical Column. 



Mercury an Evening Star. — The greatest elonga- 

 tion of Mercury (E. 25° 41') occurs on June 29. Oh 

 June 24 the planet will set at 9.54 G.M.T., or ih. 34m. 

 after the sun, and may possibly be detected close to 

 the W.N.W. horizon by anyone with a good eye. A 

 field-glass should render the planet distinctly visible 

 about an hour after sunset. 



The Zeeman Effect in Furnace Spectra. — In con- 

 tinuation of his well-known researches on furnace 

 spectra, Mr. A. S. King has recently been investi- 

 gating the Zeeman effect for iron and vanadium in 

 the electric furnace. Observations of the effect of a 

 magnetic field on spectra have up to the present been 

 chiefly confined to spark- spectra, so that it is inter- 

 esting to compare the effects when different sources 

 are used. The electric furnace possesses, in addition, 

 certain advantages over spark spectra for this pur- 

 pose, since most of the low temperature lines are 

 much more readily examined. Also, the inverse 

 effect for absorption spectra is easily obtained by 

 introducing a graphite plug to give a background of 

 continuous s[>ectrum. A description of the apparatus 

 and results is given in the Astrophysical Journal for 

 March. The furnace tube was placed parallel to 

 the lines of force in a field varving from 6500 gauss 

 in the centre to 9000 gauss near the ends, and obser- 

 vations were made of one hundred iron lines and 

 ninety vanadium lines. The results of these pre- 

 liminary observations seem to show that the effect is 

 independent of the source used, since the observed 

 separations agree both in character and magnitude 

 with those of corresponding spark lines. 



The Lunar Parallax and Related Constants. — 

 There is a set of quantities (the radius and figure of 

 the earth, the intensity of gravity, the moon's paral- 

 lax and the motion of her perigee and node) which 

 are so intimately related that an alteration in one 

 compels corresponding alterations throughout. Prof. 

 W. de Sitter has endeavoured to obtain a mutually 

 consistent series of values, and gives the results in 

 vol. xvii. of the Proc. of the Royal Academy of 

 Science, Amsterdam. It is impossible in a brief note 

 to do more than give his conclusions. 



Mean radius of earth — i.e. radius in geogr. lat. 

 the sine of which is {5 1^ = 6,37 1,237 metres. 



Value of gravity at that latitude (unaffected by 

 centrifugal force), 982014. 



ff' = sine moon's parallax/sine 1' = 3422-544". 



Compression of earth, 1/296-0. 



Constant of precession, 50-250": luni-solar pre- 

 cession, 50-373". Mass of moon, 1/81-50 

 ^^~r- —0-0032775 ; where C, A are the principal 



moments of inertia of the earth. 



•Mso, if A', B', C are the three principal moments 

 of inertia for the moon, and 



C'-B' a C'-A' B'-A' 



A' 



B' 



C 



^ is found to be 0*000626, and/=^=o-92. 

 p 



This value of / is much larger than those previously 

 found, which ranged from 0-49 to 0-75. In other 

 words, the present paper makes the moon's equator 

 less elongated towards the earth than previous deter- 

 minations. 



Prof, de Sitter's investigation reminds us of the 

 late Prof. Harkness's "solar parallax and related 

 constants." By a combination of all available evi- 

 dence he deduced, nearlv thirty years ago, a value of 

 the solar parallax practically identical with that now 

 accepted. 



