SOME CONTEMPORARY AIHWNCES IN PIIVSlCS-lll 283 



(>ni> at 1215.7 whiih is llu- first liiu- of the I.\iiian series of IndroKcn. 

 The iiilerpretalioii of spectra like this is not a simple matter of (iiittiiiB 

 eleetnxles of the desired substance into the tiihe aiul asrriltiiig to if 

 all the lines which come out on the plate. It appears that impurities, 

 e\en when present in what might he considered small proportions, 

 contribute their own rays to the spectrimi in great abundance and 

 intensity. Millikan found that all the lines present in the six-ctrum 

 of the vacuinn spark between magnesium electrodes were also present 

 when ahmiinium electrodes were used, and vice versa, and finalK' 

 assigned them all to ox\gen. Lyman foimd it extremely difficult to 

 decide which lines Ix-long to hydrogen and which to helium, since the 

 spectra of glow-discharges in these gases have so many lines in common. 

 Helium has a pronounced habit of encouraging excitation of the rays 

 of whatever other gases are mixed with it, since the helium atoms 

 require so much energy to displace their valence-electrons that free 

 electrons shot into helium gas are liable to bounce harmlessly from 

 one helium atom to another until they strike and excite an atom 

 of another variety. Even if one can be sure that all the rays in a 

 spectrum belong to a single element there remains the problem of 

 assigning them to neutral or variously-ionized atoms. It is clear 

 that the completion of the spectroscopist's task is deferred by this 

 extension of it to what the Germans call the unforeseeable time. 



We return to the consideration of the lacuna in the spectrum, which 

 extends from 13.-\ up to a boundary which by the use of high vacua, 

 concave gratings, and violent excitations, has been forced from r200A 

 down to 136.-\. This wave-length 136A stands for the moment as 

 the lowest which has ever been actually measured with the ruled 

 grating; and in spite of the unexpected and fortunate adequacy of 

 the instrument down even to this point, little more can be demanded 

 from it. The reason is, that the substance on which the rulings are 

 made must eventually cease to reflect the rays on account of its own 

 looseness of texture. Being a congeries of atoms themselves separated 

 by finite distances, the metal will not behave as a continuum towards 

 waves of a length not very large compared to its own atomic spacing. 

 Below 13.A waves are not reflected. Little is known of the rate at 

 which the reflecting-power dwindles away to zero between 136A and 

 13A and this little we owe again to Holweck. He directed a beam of 

 radiation (it was a mixed beam, as was previously made clear, and the 

 wave-length-value is merely the minimum wave-length in it) against a 

 polished bronze mirror at the very oblique incidence of 73. °8; the 

 reflected beam had one-third the intensity of the incident beam at 

 wave-length 1'23A (practically the extreme wave-Jength of Millikan's 



