SS2 



NATURE 



[December 15. 1923 



The Quantum Equivalent in Photo-electric Conduction. 



? I fretjucncy * is sent tliroiij^h a cold gas 



I f H not absorb it, 8ensitise<l l>\ admixture 



,,t ;i HI ' i' tan alworh tho light; and the 



nrt'ssurr the mean tinu* hctwcfii two 



coUisioiis IS 01 WW Name order of inaj(iiitii(ie as the 

 mean hfe of the excited state of the j^as ; all those 

 spectral lines of the non-absorbing gas apju^ar which 

 have a smaller excitation energy than n*. On the 



other 



are i 



thori 



usini 



the V 



nietlii 



betwt 



h recpiire more energy than this 

 with mixtures of mercury and 

 i>f iniTcurv ;mil s'iImt N.inoiirs, 

 1 III llic j^ \iy-; '\ ' on 



,. ..(-In- witli tlic al> i he 



1 I .111 also be used III hxniu iliu series relations 

 •II ilir lines of an I'lcnunt, since it allows us to 

 dcteniiiiir wliK li spectral liiu-s can be excited by an 

 amount oi (niii;\ smaller than a given amount. 

 Kesults lia\r Ik (11 olitained at Gottingen for lead and 

 bismuth, uhuli will shortly be ptiblished. 



In a paper in the Zeitsrlnift fiir I'hv ik, 17-3, 

 August 23, p. 202, Messrs. (i. ( aiio and J. Franck 

 consider tlic theory mon- (losdx . and dcx ribr exprri- 

 iiicnt> uhuli agree \\i\\\ ilicir ( dik hisKuis. 'I he 

 sensitisiug gas ./ has, as tlit- longest w.uc of its 

 absorption sent s, a line of frequency i', while the 

 corresponding hue of the fluorescing gas has the 

 frecniency >'i. Suppose /!•■ : ln■^ : : i : J ; then if light 

 of frequency «< is used, butli y and v^ will appear. 

 The elementary act of transmission of energy from 

 an excitcMl atom of A to an atom of B will take place in 

 siu h a \\ i\ that ] /;>• is converted into energy of transla- 

 tion of the (ohHhng atoms. If the temperature is so 

 low tiiat the kinetic energy of temperature movement 

 is small, compared with hv, the atom receives, 

 besides its excitation energy, the kinetic energy 



4 »M,i',* = "^ X ■ • Tliis abnormal velocity of 



* * * 4 I -\mi/m 



the excited atom of B produces a Doppler effect ; the 



effective frequency is Vj =1-1(1 -fcos <t>vjc), and this 



is not absorbed by the other practically resting 



atoms of B. 



An experiment with sodium vapour, sensitised with 



pure argon, has verified the theory'. Sodium vapour 



can be I mumus of the zinc i '• 



3303 A, Uo be cxtiled <lir"-. • •. 



by the If line. « oiuinion^ urreso adjiiste<l that ':.'■ 

 tul>e appeared equallv bright with tither of ih- <• 

 sources, and the light '- " tluores( mg sfwhuni m 

 each case, was passe > an absorption tuK'' 



containing soflitn" -' ,, ,.!.Ic v ' . 



light excite<l l>\ ne was 



guished. while ti it-d with ;.. ,... 



nine 1 1 less weaken 



If, in a seconil , licnt, light of freq..,.. , 

 passes t iir nnxture of gases A and li. an 



emissini ■ i can result only if the temperature 



"f ' high that '^'V difference A* 4. 



wh: :■ for th' ,n of f, ran be 



obt. ion fro; • the 



atoi. .a-nts ha\ ■;-.-i 



and mercury, and with cadmium .1 ; 



the Hg line 2536-7 A. The quar; 

 ing the vapours was pLu cd in an eleclnc uvxn. 

 which could be raised to >,oo ( ; and strong fluor- 

 was obtained wiih thalhum. In this case it 

 -ible that a thallium line, the excitation encrgv 

 o. uiiicli amounts to 5-5 volts, while the energy of 

 25367 A corresponds to 49 volts only, was due to 

 a double or step-by step process of excitation. .\t 

 800° C. part of the atoms are no longer in the normal 

 state ; and apparently there is a i v^olt excitation 

 step, from which, up to the 3-5 \olts stage, only 4-5 

 volts would be reciuired, or more than for 2536-7 A. 



With I admium and mcrciir\- this difficulty does not 

 arise, ,md the confirmation of the theory is direct. 

 It is possilile, ho\\e\er, to draw conclusions from the 

 relative intensity of the lines in the fluorescent 

 spectrum of thallium, as compared with the ordinary 

 spectrum of this substance. The differences are 

 ascribed to the differences in absorption, due to the 

 presence or absence of the Doppler effect previously 

 described ; and the combined results of the two 

 experiments seem to prove, conclusivelv, that 

 quantum energy and translation energfv can work 

 together, as an elementary act, to produce excitation 

 of the atom. 



Early Methods of Oil Painting.^ 

 Bv Prof. A. P. Laurie. 



TT is evident both from the manuscript of Theo- 

 ^ philus and the manuscript of Eraclius that the 

 properties of such drying oils as linseed oil and walnut 

 oil were thoroughly understood as early as the 12th 

 century, if not earlier. 



The methods used in their preparation differ very 

 little from the best practice of to-day. The refining 

 and bleaching of the oil and the use of driers was well 

 understood, nor is there any indication, in passing from 

 those earlier recipes to those of the 15 th century, that* 

 any new discovery of importance was made at the 

 time of the brothers Van Eyck. Passing to later 

 times, Vasari directs that pigments are to be ground in 

 walnut oil or linseed oil, and this is all that is neces- 

 sary ; he recommends the use of walnut oil as less 

 liable to darken with time. 



Many recipes for varnishes are given, and, as neither 

 spirit of turpentine nor alcohol was available in 

 commercial quantities until the end of the 15th 

 century, these varnishes are what we should now 

 describe as oil \ arnishcs, consisting of resins dissolved 

 in hot oil. The natural balsams of the pine, resin, 

 mastic, and sandarac, often all mi.xed together, were 



' Synopsis of lecture delivered at the Royal .\cademy of Arts, London, 

 on Wednesday, November 14. 



NO. 2824, VOL 112] 



used in the preparation of these varnishes. The pro- 

 portion of resinous material to oil being very high, the 

 varnishes were consequentlv vcrv stickv and had to be 

 heated and rubbed on with the hand. Spirit varnishes 

 corresponding to the mastic varni.^hes of to-dav are 

 found in i6th century and lati 



The evidence of the account- and 



Westminster show tliat lx)th oil and varnish were 

 used m pamlins,' on walls during the 13th and 14th 

 centuries, this beuiu the northern tradition, w-hile the 

 Italian tradition was the use of egg as a medium. 

 There is no indication in 1' nes of any special 



secret differing from what to-dav. 



These early pictures were p aiuLeii on a wood panel 

 sometimes covered with strips of linen and coated 

 with a gesso made of parchment size and whitening or 

 plaster of Paris which hatl been soaked in water until 

 it lost its bindimj properties. Recent e,xf)eriments 

 carried out by .Mr. Thompson at the Heriot-Wa.tt 

 College, Edinburgh, on an old i6th century panel have 

 revealed the fact that this panel was coated with a 

 non-absorbent gesso upon which a very thin layer of 

 absorbent gesso was laid so as to ensure the binding of 

 the oil to the surface of the gesso and, at the same 

 time, to preserve from staining the pure white surface 



