MOLECULES, CORPUSCLES AND IONS 283 



ous solution, is due to the late Lobry de Bruyn, and has 

 recently been verified by A. Coehn. It refers to the so- 

 called Tyndall effect. A ray of light is, as such, invisible 

 in an optically empty medium. Passed through a glass 

 trough the beam of the lantern is hardly visible, until some 

 turbid medium like smoke is introduced into the trough. 

 The light cone of a lens, concentrated into pure water, 

 leaves the water dark, when it is free of suspended particles, 

 dust, etc. It is, of course, exceedingly difficult to free the 

 water of all dust and floating impurities. Working with 

 the greatest care Lobry de Bruyn succeeded in obtaining 

 pure water, in which the light cone was hardly discern- 

 ible. But when he dissolved cane sugar in this water, 

 a luminosity was noticed. Coehn has repeated this experi- 

 ment with the ultramicroscope of Zsigmondy, and the light 

 cone then observed was quite uniform; dust particles or 

 colloids would have shown as bright points. It would, 

 therefore, appear that the large molecules of cane sugar, 

 dispersed through the water, make the water sufficiently 

 discontinuous to reflect the light. 



The further endeavors of Coehn to exemplify this dis- 

 continuity in solutions in which a transport of the ions 

 and of non-electrolytic particles, drifting with the ions, is 

 produced by electrolysis, will be better understood by a 

 description of some very remarkable experiments of Kos- 

 sonogow, of Kjew. Kossonogow studies electrolysis with 

 the aid of the ultra-microscope. He bends both the elec- 

 trodes of his cells twice at right angles, so as to leave a 

 channel, generally 0.2 millimeter in width, between the 

 active surfaces, and coats the other portions of the elec- 

 trodes with paraffin; the light beam is sent across this 

 channel in which electrolysis takes place. On dissolving 

 various salts, silver nitrate, copper sulphate, ammonium 

 chloride, and others in water, he observed at once before 

 turning on the current some luminosity and bright specks 

 in Brown i an movements. Some of the specks were no doubt 

 dust particles. But the just-mentioned discontinuity phe- 



