July 31, 1891.J 



SCIENCE. 



63 



above the melting-point. The molten metal being made the 

 negative pole, in a few hours the whole quantity had vola- 

 tilized and condensed in a thick layer on the far end of the 

 tube, near, but not touching, the positive pole. 



Volatilization of Silver. — Silver was the next metal ex- 

 perimented upon. The apparatus was similar to that used 

 for the cadmium experiments. Small lumps of pure silver 

 were cast on the ends of platinum wires, and suspended to 

 the inner ends of platinum terminals passing through the 

 glass bulb. The platinum wires were protected by glass, so 

 that only the silver balls were exposed. The whole appara- 

 tus was inclosed in a metal box lined with mica, and the 

 temperature was kept as high as the glass would allow with- 

 out softening. The apparatus was exhausted to a dark space 

 of three millimetres, and the current was kept on for 1^ 

 hours. The weights of silver, before and after the experi- 

 ment, were as follows : — 



Positive pole. Negative pole. 



Original weiglit of silver 18.14 grains. 24.63 grains. 



Weight after the experiment 18.13 " 84.44 " 



Sliver volatUlzed in 11^ hours 0.01 " 0.19 " 



In this tube it was not easy to observe the spectrum of the 

 negative pole, owing to the rapid manner in which the de- 

 posit obscured the glass. A special tube was therefore de- 

 vised, of the following character. A silver rod was attached 

 to the platiaum pole at one end of the tube, and the alumi- 

 num positive pole was at the side. The end of the tube op- 

 posite thei "ver pole was rounded, and the spectroscope was 

 arranged to observe the light of the volatilizing silver ' ' end 

 on." In this way the deposit of silver offered no obstruction 

 to the light, as none was deposited except on the sides of the 

 tube surrounding the silver. At a vacuum giving a dark 

 space of about three millimetres from the silver, a greenish- 

 white glow was seen to surround the metal. This glow gave 

 a very brilliant spectrum. The spark from silver poles in 

 air was brought into the same field of view as the vacuum 

 glow, by means of a right-angled prism attached to the 

 spectroscope, and the two spectra were compared. The two 

 strong green lines of silver were visible in each spectrum. 

 The measurements taken of their wave-lengths were 3,344 

 and 3,675, numbers which are so close to Thalen's numbers 

 as to leave no doubt that they are silver lines. At a pressure 

 giving a dark space of two millimetres the spectrum was 

 very bright, and consisted chiefly of the two green lines and 

 the red and green hydrogen lines. The intercalation of a 

 Leyden jar into the circuit does not materially increase the 

 brilliancy of the lines, but it brings out the well-known air 

 lines. At this pressure not much silver flies off from the 

 pole. At a higher vacuum the luminosity round the silver 

 pole gets less and the green lines vanish. At an exhaustion 

 of about one-millionth of an atmosphere the luminosity is 

 feeble, the silver pole has exactly the appearance of being 

 red-hot, and the volatilization of the metal proceeds rapidly. 



Like the action producing volatilization, the "red heat" 

 is confined to the superficial layers of molecules only. The 

 metal instantly assumes, or loses, the appearance of red heat 

 the moment the current is turned on or off, showing that, if 

 the appearance is really due to a rise in temperature, it does 

 not penetrate much below the surface. The extra activity of 

 the metallic molecules necessary to volatilize them is, in 

 these experiments, confined to the surface only, or the whole 

 mass would evaporate at once, as when a metallic wire is 

 deflagrated by the discharge of a powerful Leyden jar. 

 When this extra activity is produced by artificial heat, one 

 of the effects is the emission of red light; so it is not unrea- 

 sonable to imagine that when the extra activity is produced 



by electricity the emission of red light should also accom- 

 pany the separation of molecules from the mass. In com- 

 parison with electricity, heat is a wasteful agent foi" promot- 

 ing volatilization, as the whole mass must be raised to the 

 requisite temperature to produce a surface action merely; 

 whereas the action of electrification does not appear to pene- 

 trate much below the surface. 



If, for the negative electrode, instead of a pm-e metal such 

 as cadmium or silver, an alloy was used, the different com- 

 ponents might be shot off to differen distances, and in this 

 way make an electrical separation — a sort of fractional dis- 

 tillation. A negative terminal was formed of clean brass, 

 and submitted to the electrical di«charge i'w vacuo. The 

 deposit obtained was of the color of brass throughout, and 

 on treating the deposit chemically I could detect no separa- 

 tion of its component metals, copper and zinc. 



A remarkable alloy of gold and aluminum, of a rich pur- 

 ple color, has been kindly sent me by Professor Roberts- Aus- 

 ten. Gold being very volatile in the vacuum tube, and 

 aluminum almost fixed, this alloy was likely to give differ- 

 ent results from those yielded by brass, where both constitu- 

 ents fly off with almost equal readiness. The Au-Al alloy 

 had been cast in a clay tube, in the form of a rod two centi- 

 metres long and about two millimetres in diameter. It was 

 sealed in a vacuum tube as the negative pole, an alumiaum 

 pole being at the other side. Part of the alloy, where it 

 joined the platinum wire passing through the glass, was 

 closely surrounded with a narrow glass tube. A clean glass 

 plate was supported about three millimetres from the rod of 

 alloy. After good exhaustion the induction current was 

 passed, the alloy being kept negative. Volatilization was 

 very slight, but at the end of half an hour a faint purple 

 deposit was seen both on the glass plate and on the walls of 

 the tube. On removing the rod from the apparatus it was 

 seen that the portion which had been covered by the small 

 glass tube retained its original purple appearance, while the 

 part that had been exposed to electrical action had changed 

 to the dull white color of aluminum. Examined under tke 

 microscope, the whitened surface of the Austen alloy was 

 seen to be pitted irregularly, with no trace of crystalline ap- 

 pearance. 



This experiment shows that, from an alloy of gold aad 

 aluminum, the gold is the first to volatilize under electrical 

 influence, the aluminum being left behind. The purple 

 color of the deposit on glass is probably due to finely-divided 

 metallic gold. The first deposit from a negative pole (»f 

 pure gold is pink; this changes to purple as the thickness 

 increases. The purple then turns to green, which gets 

 darker and darker until the metallic lustre of polished gold 

 appears. 



If we take several liquids of different boiling-points, put 

 them under the same pressure, and apply the same amouat. 

 of heat to each, the quantity passing from the liquid to the 

 gaseous state will differ widely in each case. 



It was interesting to try a parallel experiment with met- 

 als, to find their comparative volatility under the same con- 

 ditions of temperature, pressure, and electrical iufluence. It 

 was necessary to fix upon one metal as a standard of com- 

 parison, and for this purpose I selected gold, its electrical 

 volatility being great, and it being easy to prepare in a pure 

 state. 



An apparatus was made that was practically a vacuum 

 tube with four negative poles at one end and one positive 

 pole at the other. By a revolving commutator I was able to 

 make electrical connection with each of the four negative 



