22 



SCIENCE. 



[N. S. Vol. XII. No. 288. 



If tlie chemical studj' of these metals 

 supports the conception of their elementary- 

 nature, an examination of the spectrum of 

 nickel and of cobalt, and particularly of 

 iron, forces upon us the thought of the com- 

 plexity of the atom. If each line of the 

 spectrum, representing the vibration of a 

 certain wave length, is occasioned by a cor- 

 responding vibration of the atom, it becomes 

 difficult for us to conceive of so many hun- 

 dred simultaneous vibrations of a simple 

 atom, a great share of which stand in no 

 apparent harmonic relation to each other. 

 It has been suggested that it is by a study 

 of the spectroscopic portrayal of atomic 

 vibration we may hope to gain the most 

 complete knowledge of the dynamical char- 

 acter of the atom, but it must be remembered 

 that with the spectroscope we study the mo- 

 tion of the atom at a high temperature, when 

 vibration apparently overcomes in most in- 

 stances chemical affinity ; a knowledge of 

 the atom at this temperature may give us no 

 hint whatever as to the nature of the atom 

 at lower temperatures, even as the spectrum 

 of the same element may change with vary- 

 ing temperature. "While we might thus 

 conjecture that perhaps by some process of 

 careful and refined fractionation, it might 

 be possible to resolve iron into a series of 

 meta-elements with nearly the same atomic 

 weight, we are met by the fact that com- 

 plex though it is, we find not only the same 

 spectrum for iron whatever its terrestrial 

 source, but that the spectrum of sidereal 

 iron, from meteorite, from sun, from star, 

 gives us no evidence of any variation in the 

 composition of iron. We are, I think justi- 

 fied in concluding that the nine metals of 

 the eighth group fulfill every definition of 

 an element, and that they are just as much 

 to be looked upon as simple elementary 

 substances as any of those substances which 

 we call elements ; and further that while 

 refined determinations may change, to a 

 Slight extent, the atomic weight of some of 



these elements, especially those of ruthen- 

 ium and osmium, we may expect the weight 

 of these elements relative to each other, 

 and hence their position in the periodic 

 sj'stem to remain unchanged. This, of 

 course, carries with it the conclusion that 

 in the periodic table an element may have 

 an atomic weight slightly lower than that 

 of the element which precedes it. I have 

 discussed this possibility briefly elsewhere,* 

 and will only add that seeming exceptions to 

 accepted laws, instead of overthrowing the 

 law, often serve to broaden our conception 

 of the law itself. 



Before considering some of the com- 

 pounds of the metals of the eighth group, 

 attention must be called to the phenomenon 

 exhibited by several of these metals, and par- 

 ticularly by palladium, of condensing hy- 

 drogen and other gases upon their surface. 

 The first observation in this connection 

 seems to have been that of Sir Humphrey 

 Davyf, who in 1817 showed the Eoyal 

 Society how a warm platinum wire, plunged 

 into the vapor of alcohol or ether or cer- 

 tain other inflammable gases, became incan- 

 descent, and continued to glow as long as 

 kept in the vapor, causing an oxidation of 

 the gas and in some mixtures even an ex- 

 plosion. This phenomenon attracted great 

 attention on the part of chemists and many 

 were the discussions over this lamp ' with- 

 out a flame ' or Davy's ' aphlogistic lamp ' 

 as it was called. It was soon after noticed 

 by Edmund Davy % that the platinum re- 

 duced from solution, now called platinum 

 black, but then platinum suboxid, is espe- 

 cially active and can oxidize alcohol to 

 acetic acid. In 1823 Dobereiner announcedg 

 that platinum black and platinum sponge, 

 when held in a stream of hj'drogen, ignite 

 the gas and that the hydrogen is absorbed 



*Chem. News, 80, 74 (1899). 



\PhiI. Trans., 107, 77 (1617). 



tibid., 110, 108 (1820). 



? J", fiir. Chem. (Schweigger), 38, 321 (1823). 



