TllANSACTIONS OF SECTION A. 455 



spectra of the same substance at different temperatures, Dr. Iluggins gives us the 



answer. ,. , , ^. . t i 



•^ In the stars which give a very tvhite hght, such as binus or a Lyrse, we have 

 the lines G and h of hydrogen and also 11, which has been lately shown by Dr. 

 Vc.o-el to be coincident with a line of hydrogen ; but the K line of calcium is weak 

 in a Lyrfo, and does not appear in Sirius. In passing from the white or hottest 

 stars to the yellow stars like our sun, the typical lines diminish in breadth and are 

 better defined, and K becomes stronger relatively to H, and other lines appear. In 

 \rcturus we have a star which is probably cooler than our sun, and m it the line K 

 is stronger in relation to H than it is in the solar spectrum, both being very strong 

 compared with their state in the solar spectrum. tt • i 



Professors Liveing and Dewar find that K is more easily reversed than H in the 

 electric arc, which agrees with the idea that this line is produced at a lower 

 temperature than H. , i , v 



Besides the absence or weakness of K, the white stars have twelve strong lines 

 winged at the edges, in which there are three of hydrogen, ^az. G, /(, and H, and 

 the "remaining nine form a group which are so related to one another that Dr. 

 Huggins concludes they probably belong to one substance. Three of these Imes are 

 said by Dr. Vogel to be" lines of hydrogen. . . 



Professors Liveing and Dewar have made considerable progress m determining 

 the conditions and the order of reversal of the spectral lines of metallic vapours. 

 They have adopted methods which allow them to observe through greater thiclc- 

 nesses of vapour than previous observers have generally employed. For lower 

 temperatures tubes of iron or other material placed vertically m a furnace were 

 used, and the hot bottom of the tube was the source of light, the absorption being 

 produced by vapours of metals dropped into the hot tube and filling it to a greater 

 or less height. By this means many of the more volatile metals, such as sodium, 

 thalhum, nidium, ciBsium, and rubidium, magnesium, lithium, barium, strontium, 

 and calcium, each gave a reversal of its most characteristic line or pair of lines, 

 i.e. the red line of lithium, the violet lines of rubidium and calcium, the blue line 

 of strontium, the sharp green line of barium (5535), and no other lines which can 

 certainly be ascribed to those metals in the elementary state. 



For "higher temperatures tubes bored out in blocks of lime or of gas carbon, 

 and heated by the electric arc, were used. By keeping up a supply of metal and 

 in some cases" assisting its volatilisation by the admixture of a more volatile metal, 

 such as mao-nesium, and its reduction by some easily oxidisable metal, sucJi as 

 ahimmiiun, or by a current of coal gas or hydrogen, they succeeded m maintauiing- 

 a stream of vapour through the tube so as to reverse a great many lines. In this 

 way the greater part of the bright lines of the metals of the alkalies and alkahne 

 earths were reversed, as well as some of the strongest lines of manganese, alu- 

 minium, zinc, cadmium, silver, copper, bismuth, and the two characteristic lines of 

 iridium and of gallium. By passing an iron wire into the arc through a perforated 

 carbon electrode they succeeded in obtaining the reversal of many of the lines 

 of iron. In observing bright line spectra they have found that the arc produced 

 by a De Meritens machine arranged for high tenoion gives, in an atmosphere ol 

 hydrogen, the lines C and F, although the arc of a powerful Siemens machine doe.-* 

 not bring them out, and they have observed many metallic lines in the arc which 

 had not been previously noticed. The temperature obtained by the De Meritens 

 machine is thus liigher "than that obtained in the Siemens machine. 



From observations on weighed quantities of sodium, alone and as an amalgam, 

 introduced into a hot bottle of platinum filled with nitrogen, of which the pressure 

 was varied by an au--pump, they conclude that the width of the sodium lines 

 depends rather on the thickness and temperature of the vapour than upon the 

 whole quantity of sodium present. Very minute quantities diffused into the cool 

 part of the tube give a broad diffuse absorption, while a thin layer of compressed 

 vapour in the hot part of the tube give only narrow absorption lines. Professors 

 Livemo- and Dewar have observed the reversal of some of the well-known bands of 

 the oxides and chlorides of the alkaline earth metals. The lines produced by 

 mai-nesium in hydrogen form a rhythmical series extending all across the well- 



