46 



NATURE 



{May 8, 1879 



degree according to the material nsed, on the upper and lower 

 limits between which, since the last complete release from ten- 

 sion, the tensions have varied (the succession alone, not the time, 

 being here determinant). In a thermo element of two similar 

 wires, sti etched with equal weight:, but of which one (a) has 

 last borne a greater, the other {b) a less weight, the current flows 

 from b to a. In iron and steel the previous tensions even affect 

 the direction in which the thermo-current varies with increasing 

 or decreasing tension. Taking the direction of the current which 

 arises with theyfrj/ weak stretching, it is opposite for hard and 

 soft wires, and the discordance above referred to is thus ex- 

 plained. Herr Cohn thus formulates his general conclusion : "A 

 stretched wire behaves, ceteris paribus, differently, according as 

 it has before been stretched more weakly or more strongly, and 

 this difference continues till the next alteration of the tension." 

 — Herr R. Weber contributes a usefiil paper on the chemical 

 composition of glasses with relation to their resistance to atmo- 

 spheric influences. He finds that the composition of many well 

 tested lime alkali glasses approximates the proportions 6Si05, 

 iCaO, and iKjO or Na^O ; but also, in good glasses, there may 

 be more alkali, if it be compensated with more than 6 equiva- 

 lents of silicic acid ; and less silicic acid may be allowed if the 

 lime be diminished relatively to the alkali. — The transmission of 

 high tones through the telephone is discussed by Herr Hagen- 

 bach. From his experiments it appears that the (upper) limit of 

 audibility with the instrument is commonly about two octaves 

 lower than in direct hearing. The cause is found not in the line, 

 nor in the magnet, but in the plate, which, when the variations 

 of magnetism exceed a certain number per second, no longer 

 keeps up with them. — Herr Aron gives a mathematical study of 

 the microphone ; inter alia, it is shown that, whereas in the tele- 

 phone the "clang tint" is exalted, in the microphone it is 

 lowered. — Herr Herwig prosecutes his study of liquid cells as 

 condensers ; considering the charge of cells, first by large con- 

 stant batteries, then by small forces (both acting shortly), and 

 comparing the full charges in cells containing liquids of different 

 resistances. — We farther note a new hygrometer by Herr 

 Edelmann, based on the fact that if any space be filled with at- 

 mospheric air, and the aqueous vapour then removed, without 

 altering the volume, the pressure decreases by the amount of ten- 

 sion of this vapour. A sinus manometer for measuring small dif- 

 ferences of air-pressure (Thomsen), and a simple regulator for the 

 electric light (Stohrer) are also described. — Herr Frohlich shows 

 the bearing of the principle of conservation of energy on the 

 theory of diffraction. 



The Rivista Scientificolndustriale, (1879, No. 7) contains the 

 following papers of interest : — On some prehistoric discoveries 

 in Sicily, by Prof. Francesco Mangini. — On the lengthening of 

 filiform conductors traversed by an electric current, by Prof. G. 

 Basso. — On the optic rotatory power of quartz and its variation 

 with temperature, by M. Joubert. — On digallic acid, by Prof. 

 U. Schiff. — On some phenomena due to the viscosity of liquids, 

 by Prof. Carlo Marangoni. — On Sargasso seas, by the same. — 

 On Helmholtz' double siren, by the editor. 



SOCIETIES AND ACADEMIES 

 London 



Royal Society, February 20 and 27, and Aprils. — "On ^^ 

 Reversal of the Lines of Metallic Vapours," Nos. iv., v., and vi., 

 by Professors Liveing and Dewar. 



In the first of these experiments the metals were produced by 

 chemical reactions within the tubes, used as before described. 

 They found that caesium chloride alone heated in glass tubes gave 

 no absorption lines, but ca;sium and rubidium chlorides when 

 heated with metallic lithium each showed its characteristic 

 absorption lines. Charred tartrate of caesium heated in a furnace 

 in a narrow porcelain tube gave very readily the two lines in the 

 blue reversed, and charred rubidium tartrate the two violet lines 

 rcTcrsed, but no reversal in any other part of the spectrum. 

 When charred potassium tartrate was treated in the same way, a 

 broad ab^o^ption band was seen extending from wave-length 

 about 5,700 to about 5,775. This band was also seen for a short 

 time bright, when the material was put into the tube before it was 

 heated, and the light observed as the tube got hot. It was also 

 seen bright in the induclion-spark taken between platinum and 

 potassium in carbonic oxide. Besides this band the vapours from 

 the charred potassium tartrate produced another absorption band 

 in the red, and two more in the blue. None of these absorp- 



tions 'correspond with those seenwhen potassium is heated in 

 hydrogen, or with known emission lines of that metal, though 

 the first and most conspicuous is near a well-known group of 

 three bright lines of potassium. Charred sugar mixed with 

 carbonate of soda gave only the same absorption as sodium in 

 hydrogen. A mixture of barium carbonate, lamp-black, and 

 aluminium filings gave dark bands corresponding to the bright 

 bands seen when sparks are taken from a solution of barium 

 chloride marked o, fl, and 8 respectively by Boisbaudran ; and 

 at the highest temperature of the furnace fed with Welsh coal a 

 mixture of charred barium tartrate and aluminium gave the 

 barium line wave-length 5535 sharply reversed. Charred stron- 

 tium and calcium tartrates with aluminium gave no reversals, but 

 with the addition of sodium or potassium carbonate the well- 

 known blue line of strontium and violet line ol calcium were 

 reversed. The temperature at which these results were obtained 

 was reached by the use of gas retort carbon as fuel, and was such 

 that iron tubes well coated with fire-clay gave way in a few 

 minutes. 



The next experiments were made with tubes bored out of lime 

 and heated at the bottom by a jet of coal-gas and oxygen intro- 

 duced through a lateral opening. In these, as in the previous 

 experiments, the hot bottom of the tube itself (not an indepen- 

 dent light as used by Messrs. Lockyer and Roberts in their 

 experiments with lime tubes) gave the luminous background. 

 In this way the violet line of calcium was reversed, the red line 

 of lithium and the orange and green bands of lime appeared 

 with dark lines down their middles. 



A larger series of experiments was made with similar tubes of 

 lime, but with an electric arc introduced through lateral open- 

 ings as the source of light and heat. In some cases a tube 

 bored in a block of gas carbon was :employed, and was then 

 made one of the electrodes. The carbon tubes, however, were 

 found to conduct away the heat, and though they lasted much 

 longer, did not in general succeed so well as the lime tubes. In 

 same cases aluminium was used as a reducing agent, and in 

 others more volatile sitbstances, viz., potassium and sodium car- 

 bonates, were used to increase the amount of vapour carried up 

 into the tube ; and in others a current of hydrogen was intro- 

 duced. 



Of the calcium lines the violet line (4226) was almost always 

 seen expanded with a dark middle, and the three brightest lines 

 in the indigo were often in a similar condition. The addition of 

 aluminium generally made them appear as dark bands on a con- 

 tinuous background. Of the two Fraunhofer lines H, the more 

 refrangible (K) was the first to appear reversed, and remained so 

 the longer. Other calcium lines reversed were, one in the green 

 (5188), and, much less easily, two in the red (6161, 6121), one 

 more in the indigo {4302), and one in the blue (4877). 



In the case of strontium, the well-known blue line was easily 

 reversed, and two lines in the violet {4215, 4077), less easily five 

 lines in the blue (4812, 4831, 4868, 4873, 4895), and, by the 

 aid of aluminium, one in the green (4962). In the case of 

 barium, besides the persistent ray 5535, two other lines in the 

 green (5518, 4933), a line in the blue (4553), and one in the 

 orange (6496), were reversed. 



With magnesium the b group were expanded and reversed in 

 an order the inverse of their refrangibility. The other lines of 

 that metal were expanded, but not reversed, and the blue line, 

 4481, conspicuous in the spark between magnesium electrodes, 

 was not seen at all. This line does not appear in Capron's pho- 

 tographs of magnesium in arc. An attempt to re-introduce it by 

 combining the action of an induction spark with that of the arc 

 in a lime tube failed owing to the conducting power of the ;hot 

 gases and walls of the tube, and will probably only succeed with 

 a pressure of several atmospheres in the apparatus. The similar 

 disappearance of the cadmium lines 5377 and 5336 was also 

 noticed. 



In using potassium carbonate 'the two extreme pairs of lines, 

 in the violet and red respectively, were readily reversed ; less 

 readily the three lines in the greenish-yellow, other two lines in 

 the red (6913, 6946), a group of three in the orange (5353, S338, 

 5322), and the least refrangible (5 11 2) of another triplet in the 

 green. 



Using sodium chloride, the pair of lines next more refrangible 

 than D were rej-ieatedly reversed, the less refrangible being the 

 first and most strongly reversed, as has also been observed by 

 Mr. Lockyer. A second pair of bright lines usually came out at 

 the same time, like ghosts of the first, on the more refrangible 

 side. 



With lithium chloride, the red and blue liner were easily rt- 



