140 flF.LL SYSTEM TF.CIIXIC.IL .lOL'RX.-II. 



iii}i n(.Kativul\ -rhargcd wiit-s or plates into ihc i>usili\e column. These 

 wires and jilales surroundeil themselves with dark sheaths, the thick- 

 ness of which increased as the potential of the metal was made more 

 and more hii;hly negative. The explanation is, that the electrons 

 in the positive column cannot ai)i)i-oach the intruded wire, being 

 clri\eii back b\- the adverse field; the ilark sheath is the region from 

 which they are excluded, and across it the positive ions advance to 

 the wire through a field controlled by their space-charge. The equation 

 selected by Langniuir to represent the relation between the thickness of 

 the sheath, the voltage across it, and the current of positive ions into 

 it, is (16). As the sheath is \ isible and its thickness can be measured, 

 as well as the other cjuantities, the relation can be tested. This was 

 done by Schottk)-; the result was satisfactory. When the intruded 

 electroile is a wire, the sheath is c>lindrical, and expands as the voltage 

 of the wire is made more negati\e. As the area of the outer boundary 

 of the sheath is increased by this expansion, more ions from the positi\'e 

 column touch it and are sucked in, and the density of flow of positive 

 ifins in the column can be determined. By lowering the potential 

 of the wire gradually so that the electrons can reach it, first the fastest 

 and then the slower ones, the velocity-distribution of the electrons 

 in the column can be ascertained. Their average energy depends on 

 the density of the mercury vapour, and may amount to several \-olts. 



Beyond the posili\'e colimin lies the anode, itself preceded by a 

 sharp and sudden potential risi-. The electrons are flung against it 

 with some force, and it grows and remains \ery hot; usually, in fact, 

 hotter than the cathode. This high temperature does not seem to 

 be essential to the continuance of the discharge, for the anfxle can 

 be c<Hjled without killing the arc; yet it seems strange that a quality 

 so regularly found should be without influence upon th? discharge. 

 One must beware of underestimating the influence of the anode; 

 when an arc is formed in air between two electrodes of different 

 materials, it behaves like an arc formed between two electrodes of 

 the same material as the anf)de, not the cathode! 



The so-called lo'u'-i'ollanc arc, although not a self-niaini, lining dis- 

 charge, merits at least a paragraph. .A dense electron-sireani poured 

 into a monatomic gas from an independenth-hcaleil wirr. and ac- 

 celerated by a P.D. surpassing the resonance-potential of liiu gas, 

 may ionize it so intenseh- that there is a sudden transformation into 

 a luminous arc-like discharge. This is a sort of "assisted" arc, its 

 catlxKle being ke|)t warm for its benefit by outside agencies. Its 

 history is .i long and interesting chapter of contemporary physics, 

 whereof the end i> not \et. The most remarkable feature of this arc 



