924 Sir James Deivar [Jan. 21, 



shows an even greater deflection, which does not die away so quickly 

 unless we short circuit the cell. Such electrical effects depend upon 

 the temperature of the cell. On cooling the uranium nitrate cell 

 with a pad of cotton wool saturated with liquid air the alcohol will 

 be cooled down to its freezing point, and now no electrical action 

 takes place. In fact a temperature of only - 80° C. is sufficient to 

 arrest any current, the solution being then congealed to a jelly. 



The ordinary photographic action of light is similarly prevented 

 by cooHng to a low temperature. A piece of photographic paper 

 kept cooled locally in a patch of about 3 in. diameter by a cotton 

 wool pad soaked in liquid air while exposed to the electric beam is 

 blackened, except in the cooled patch which remains unacted upon. 



The photographic action of ordinary light is largely restricted to 

 the ordinary temperature. When we come to ultra-violet Hght we 

 have a different state of things. This radiation is capable of pro- 

 ducing effects at the temperature of liquid air and even liquid hydrogen. 

 A convenient and powerful source of ultra-violet light is the electric 

 arc in mercury vapour contained in silica tubes. Glass tubes would 

 of course absorb the ultra-violet light besides being liable to fracture 

 from the high temperature of the lamp. An arc between copper and 

 carbon poles in air is also available as a source of ultra-violet light, 

 but is not so convenient to work as a modern mercury vapour lamp. 

 In all my later experiments on phosphorescent bacteria the mercury 

 lamp has been used. The general arrangement of the apparatus is 

 shown in Plate II. figs. i. and ii. Liquid oxygen is highly absorptive 

 of the ultra-violet hght. If a few cubic centimetres of hquid oxygen 

 are poured into a shallow silver dish floating on liquid air, and a 

 quartz cell containing water to the depth of a centimetre is arranged 

 to absorb the bulk of the heat radiation, the ultra-violet rays can 

 continue to act on the liquid oxygen for a considerable time. 

 Pouring the remaining liquid oxygen into an ordinary glass beaker, 

 in which are suspended some strips of paper treated with potassium 

 iodide and starch solution, the presence of ozone in the evaporating 

 liquid is evident from the dark blue colour produced. To prove 

 that this is the effect of the ultra-violet radiation the experiment 

 will be repeated, with a very thin lamina of mica, which is opaque to 

 the ultra-violet, interposed between the mercury lamp and the liquid 

 oxygen. In this case no ozone is produced, the strips of paper re- 

 maining quite white after the exposed oxygen is poured into the 

 beaker. The smell of ozone during the evaporation of the liquid air 

 is the most delicate and characteristic property of the body. If the 

 ozone has to be estimated quantitatively, then the liquid oxygen after 

 exposure to the ultra-violet radiation of the arc or mercury lamp is 

 transferred to a small vacuum vessel B (Plate I. fig. iv.) and 

 allowed to slowly evaporate, the gas being bubbled through iodide of 

 potassium solution in the vessel A. In order to prove that the 

 formation of ozone can take place in liquid oxygen, avoiding any 

 action on the gas, a quartz vacuum vessel (Plate I. fig. iii.) had a 



