November 24, 1904] 



NATURE 



95 



of metals. In connection with the work on enhanced lines, 

 it has been noted that some of them, at least, appear to 

 correspond with comparatively weak solar lines to which 

 Rowland has attached no origin. With the object of 

 possibly tracing some of the unorigined solar lines to their 

 source, a careful comparison has been made between the 

 enhanced lines shown in the photographic spark spectra of 

 Ti, Fe, and Cr and the solar lines. The photographs used 

 for this purpose were all taken with a Rowland grating, and 

 on such a scale that the length of spectrum between K 

 and F is about 14 inches (35 cm.). The chemical elements 

 named were first selected for investigation because they 

 furnish by far the greater number of enhanced lines which 

 have been shown 10 occur in the spectrum of a Cygni. 



It was found that many of the enhanced lines fell exactly 

 on isolated lines of the solar spectrum, and in these cases 

 the solar wave-lengths were adopted and the identification 

 considered established. If, however, for any of these solar 

 lines Rowland had given alternative origins, special com- 

 parisons were made of the enhanced line photograph with 

 those of the metals given by Rowland. Notes (given at 

 the end of the tables) were made as to the agreement or 

 non-agreement of the metallic lines involved, and also of 

 the relative intensities in their individual spectra, so that 

 due weights could be given to the respective metallic lines 

 which were thought conjointly to produce compound solar 

 lines. 



Where there was any doubt as to the e.xact coincidence 

 of a metallic and solar line, or where by the close grouping 

 of several solar lines it was not possible to say by direct 

 comparison to which solar line the metallic line corre- 

 sponded, careful measures were made of the metallic line, 

 and its wave-length found by interpolation between closely 

 adjacent lines of known wave-length. The resulting wave- 

 lengths were then compared with Rowland's solar wave- 

 lengths, and in cases of close agreement with solar lines it 

 was deemed probable that the two lines were really identical. 



A final table is given of the enhanced lines of the three 

 elements which are considered, as a result of the analysis, 

 to be identical with lines in the Fraunhoferic spectrum. 

 Forty-two of these agree with solar lines unorigined by 

 Rowland, and as the majority of them are conspicuous 

 lines in stellar spectra of certain types, it has been thought 

 that these results will be of importance in standardising 

 the wave-lengths of many stellar lines. 



Physi'-al Society, November II. — Dr. R. T. Glatebrook, 

 F.R.S., president, in the chair. — Investigation of the varia- 

 tions of magnetic hysteresis with frequency : Prof. T. R. 

 Lyie. The experiments were made on two rings of lamin- 

 ated annealed iron, in one of which the radial breadth of 

 the iron was considerable relative to its mean radius. 

 These rings were magnetised by alternating currents of 

 different strengths and periods ; both the magnetising- 

 current wave and the magnetic-flux wave were quantitatively 

 determined by a wave-tracer (described by the author in 

 the Fhil. Mag., November, 1903), and the wave-forms so 

 obtained subjected to harmonic analysis. The experiments 

 were divided into series, in which the period and wave-form 

 of the magnetising current were kept as nearly constant 

 as possible throughout any one series, while its strength 

 was varied. The analytic expressions for the associated 

 current and flux waves for a few series are given in tabular 

 form. From the analytic expressions for each pair of 

 associated waves it was found that when the magnetising 

 current was approximately sinusoidal the total iron loss 

 (I) was, within certain limits of the induction, given by a 

 formula I = (a + bH)93' ''*' where n is the number of periods 

 per sec. 33 the '"effective induction," and a and b are 

 constants. When from the total iron loss per c.c. per cycle 

 the sum of the statical hysteresis and the value that 

 theory assigns to eddy-current loss was subtracted, a con- 

 siderable quantity remained, which increased both when 

 the frequency and the flu.x-density increased. This quantity, 

 called by Fleming the kinetic hysteresis, has been obtained 

 for each experiment. — On the practical determination of the 

 mean spherical candle-power of incandescent and arc lamps : 

 G. B. Dyke. Mr. Dyke points out the need of an improved 

 method of expressing the efficiency of glow-lamps, and 

 adopts the suggestion of Dr. Fleming of expressing the 

 whole flux of light in lumens per watt. The expression of 



NO. 1830, VOL 71I 



the efficiency in this manner involves the determination of 

 the mean spherical candle-power (M.S.C.P.), and the paper 

 describes a method of doing this. The objects of the paper 

 are : — (i) to obtain curves showing the variations of candle- 

 power of glow-lamps in a horizontal plane; (2) to obtain 

 reduction factors by which the mean horizontal candle- 

 power (M.H.C.P.) may be calculated from the maximum 

 horizontal candle-power (C.P.); and (3) to obtain reduction 

 factors for deducing the M.S.C.P. from the M.H.C.P. and 

 from the C.P. — Exhibition of apparatus : R. W. Paul. The 

 construction of highly sensitive pivoted electrical instru- 

 ments has been rendered difiicult by the fact that delicate 

 pivots will not admit of transporting without injury. A 

 number of galvanometers were shown in which the design 

 was based upon the use of a moving coil, supported on one 

 pivot in a powerful and uniform magnetic field, and con- 

 trolled by a spring. A simple non-reflecting, suspended-coil 

 galvanometer for the student's use, with a sensibility of 

 I division per micro-ampere, was also exhibited. A new 

 design of lantern, adapted for science lectures, and for use 

 with three Nernst filaments arranged closely together, was 

 shown in action. It is capable of being instantly changed 

 from horizontal to vertical projection, can be fitted with a 

 reversing prism, and has a wide adjustment for focusing. 

 Another exhibit was an Ayrton Mather reflecting electro- 

 static voltmeter with a magnetic damping device. The 

 instrument shown had a sensibility of 500 mm. at i m. foi 

 30 volts, but similar instruments are made to give this 

 deflection with pressures as low as 8 volts. 



Paris. 

 Academv of Sciences, November 14. — M. Mascart in the 

 chair. — Researches on the desiccation of plants : the period 

 of vitality. Moistening by liquid water : imperfect reversi- 

 bility : M. Berthelot. — New researches on the Carion 

 Diablo meteorite : Henri Moissan. A very careful and 

 complete examination was made of a block of this meteorite 

 weighing 183 kilograms. It was found to be distinctly 

 heterogeneous in structure, containing iron, nickel, sulphur, 

 phosphorus, silicon, and carbon. The latter element was 

 present in several forms : amorphous carbon, graphite, and 

 diamonds, both the black and transparent variety of the 

 diamond being separated. Characteristic green hexagonal 

 crystals of silicon carbide were also isolated, the author re- 

 marking that this is the first time that this compound has 

 been met with in nature. — The measurements of the velocity 

 of propagation of earthquakes : G. Lippmann. An instru- 

 ment is described capable of determining to 1/5 of a second 

 the exact time of the commencement of a seismic shock at 

 any given point. The author also discusses the following 

 problem : to find the direction of the seismic wave front at 

 the surface of the earth, in a given region, and to measure 

 the velocity of its horizontal propagation. — On the inscrip- 

 tion of seismic movements ; G. Lippmann. In the photo- 

 graphic self-recording apparatus in common use for earth- 

 quake phenomena, owing to the considerable expense of the 

 strip of sensitised paper, its velocity through the apparatus 

 is very slow, about 12 cm. per hour. In the modification 

 now proposed, the slit through which the ray of light falls 

 on the paper is closed by a shutter, and this is operated 

 electrically by the seismic shock. By this means the speed 

 may be greatly increased, since the paper is only used up 

 during the period of the earthquake shocks. — On the seeds 

 of the NeuropteridccC : M. Grand'Eury. As the result of 

 the examination of more than 1000 specimens of fossil 

 seeds, usually attributed to ferns, the author distinguishes 

 15 genera or subgenera of NeuropterideiE, and 25 specific 

 types. — Remarks on Hugoniot's adiabatic law : M. 

 Jousuet. — On the use of helium as a thermometric sub- 

 stance and on its diffusion through silica : Adrien 

 Jaquerod and F. Louis Perrot. .An attempt to deter- 

 mine the melting point of gold with a thermometer of fused 

 silica, and containing helium, failed owing to the rapid 

 difl'usion of the gas through the silica at the high tempera- 

 ture. The velocity of diffusion appears to be proportional 

 to the pressure of the gas, and is very considerable, since 

 after six hours' heating at 1100° C. the pressure of the 

 helium had fallen to about one-seventh of the initial pressure. 

 Below a red heat, at about 510° C, the diffusion is still 

 fairly rapid, and a very slow effect could even be traced 

 at 220° C. For practical purposes, therefore, the nitrogen 



