6o 



NA TURE 



[Mav i8, 1893 



lished next month. It is based upon the articles contributed by 

 him to the ninth edition of the " Encyclopcedia Britannica," 

 but contains large numbers of others by himself and Dr. 

 Gadow, the Strickland Curator at Cambridge, together with 

 contributions by Mr. Lydekker, Prof. Roy, and Dr. Shufeldt. 

 The work is to consist of four parts, and when complete will 

 form a demy 8vo volume of about 1000 pages, copiously illus- 

 trated, and the publishers Messrs. Adam and Charles Black, 

 promise the second part in October next. 



In the current number of the American Journal of Science 

 Mr. Pupin gives the second part of his paper on electrical 

 oscillations of low frequency and their resonance (see Nature, 

 April 20, 1893). This portion consists chiefly of a theoretical 

 investigation of the rise of potential in a circuit which is in 

 resonance w ith a periodical impressed electromotive force. This 

 rise was shown in a very striking manner by connecting two 

 large choking coils and a condenser in series with the secondary 

 of a transformer, a Thomson electrostatic voltmetre being con- 

 nected to the terminals of the condenser, and the core of one of 

 the choking coils consisting of a movable bundle of solf iron 

 wires. The frequency of the impressed electromotive force 

 being about 100 per second, the capacity of the condenser was 

 adjusted until the removal of the plug was accompanied by 

 bright sparks, showing that resonance was near. Then the 

 movable iron core was adjusted till the voltmetre gave the 

 longest deflection obtainable. In this way a rise from 60 volts 

 (generated in the secondary and indicated by a Cardew volt- 

 metre) to about 900 volts was easily obtained. The rise of 

 potential is practically confined to the condenser, there being, 

 however, a large and rapid increase in the current on the 

 approach of resonance, which increase can be studied in a rough 

 way by the pull which the choking coil exerts on the movable 

 iron core when this is being adjusted to give resonance. 



The question as to the cause of earth currents is one of con- 

 siderable interest, and has not yet been satisfactorily answered. 

 A paper read before the Institute of Electrical Engineers by Mr. 

 O. E. Walker, giving a further account of his observations made 

 in India seems to show, however, a clear connection between 

 the variations in the earth currents and those of the atmospheric 

 pressure. On account of the long periods of settled weather 

 experienced in India, that country is particularly well suited for 

 the investigation of any such relation. Observations were made 

 on four lines, and show, in almost every case, that in the 

 morning the current flows from the inland place of observation 

 to the coast, while in the afternoon the direction is reversed. 

 The maximum current in one direction occurs at 10 a.m. local 

 time, this also being the time of the morning maximum reading 

 of the barometer, while the maximum current in the opposite 

 direction occurs at 3 p.m., the time of the afternoon minimum 

 of the barometer. Another point of intere>t is that the maxi- 

 ma of the earth currents occur when the diurnal variation of 

 the declination is zero. 



On a clear, cold, winter afternoon, about half an hour before 

 sunset, a peculiar phenomenon of refraction of light can be 

 witnessed on a fresh field of snow, which is described by Mr. 

 Albert W. Whitney in the American Journal of Science. Two 

 roughly V-shaped paths, of especial, though not exclusive 

 brilliancy, open away from the observer and towards the sun. 

 The apex of one is perhaps six feet away, its angle 90' ; the 

 apex of the other is perhaps fifteen feet away, its angle 60°. 

 The light is not diffused, but made up of many separate brilliant 

 points, glowing with prismatic colours. The paths are broad, 

 several degrees in width, their inner margin being more sharply 

 defined than their outer limit. Measurements with a sextant 

 have shown that all the glowing points lie in the surfaces of 

 two cones, whose axes pass through the sun and the eye of the i 

 NO. 1229 VOL. 48] 



observer, and whose angles are approximately 22' and 46°. 

 Hence the paths are hyperbolas, and their visibility depends 

 upon the altitude of the sun above the horizon. Mr. Whitney 

 explains the phenomenon as analogous to that of halos. It is 

 largely, perhaps mostly, due to frost crystals. They form more 

 slowly, hence more regularly, than snow crystals. The fact 

 that snow hyperbolas are usually more conspicuous in the late 

 afternoon than in the early morning, may be explained by the 

 frost crystals needing a certain amount of clearing up by sun 

 and wind of minute secondary accumulations of frost upon 

 themselves, to make them fit for transmitting light. Another 

 interesting fact concerns the relation of the other limb of the 

 hyperbola to that upon the snow. If the observer walks so as 

 always to keep one certain point in the path of light, his track 

 will be an hyperbola ; if now, from the apex of the hyperbola 

 wliich he has traced, he advances a distance equal to his height 

 multiplied by the cotangent of the altitude of the sun plus half 

 the vertical angle of the cone, the figure which he now sees and 

 the figure which he has traced upon the snow are the two limbs 

 of the same hyperbola. 



The determination of the refractive index of the atmosphere 

 has until recently been confined to the visible spectrum. Messrs. 

 Kayser and Runge, in a communication to the Berlin Academy,, 

 describe a method of obtaining it for every portion of the photo- 

 graphic spectrum. If a prism is introduced between a Rowland 

 concave grating and the sensitive plate, the rays are deflected, 

 and the spectrum appears displaced on the plate. From the 

 amount of this displacement and the distance of the prism from 

 the plate it is easy to deduce the angle of deflection of the rays in 

 question, and the refractive index of the prism. The prism used 

 was hollow, made of copper and closed by quartz plates. The 

 prism wasfilled with air under a pressure of aboutten atmospheres 

 Since the investigations of Mascart, Benoit, and Chappuis £.nd 

 Riviere have shown that the index of deviation of gases varies 

 as the density, the refractive indices for air at zero and atmo- 

 spheric pressure could be calculated from the results. Measure- 

 ments were taken at seven different places along the spectrum, 

 between wave lengths 563 and 236/1/4. The refractive indices- 

 for some of the Fraunhofer lines thus obtained are the: 

 following : — 



A I '0002902 



D I '0002919 



F I '0002940 



G I '0002959 



H I '0002975 



N I '0003000 

 X = 236/iM I '0003217 

 These values are for air under normal temperature and pressure 

 but not for dry air. To correct for moisture the last decimal 

 place must be increased by 3, thus giving forD, for instance, the 

 value 1*0002922. In the hands of previous workers, the last 

 two figures have varied between u (Lorenz) and 47 (Ketteler). 



No fewer than six volumes containing French translations of 

 writings by Mr. T. H. Huxley are now included in the " Biblio- 

 theque Scientifique Contemporaine," issued by J. B. Bailliere 

 et Fils. The latest of these six volumes is a series of essays, 

 entitled "Science et Religion." 



A volume embodying the meteorological results deduced from 

 observations taken at the Liverpool Observatory during the 

 years 1889-90 91 has been published by order of the Mersey 

 Docks and Harbour Board. 



The atomic refraction of nitrogen, in the free state as gas and 

 in the various types of nitrogen compounds, forms the subject of 

 a communication to the Berichte by Prof. Briihl, of Heidelberg. 

 The refractive power of free gaseous nitrogen has been deter- 

 mined with the highest attainable accuracy by Biot and Arago, 

 Dulong, and other later observers, and the value of the atomic 



P 



; and pressu 



