PHOTOCHEMISTRY 279 



It is necessary to measure the energy absorbed in the whole reaction 

 vessel. The vessel is set up in the optical train back of the mono- 

 chromator slit or back of a filter and the intensity of the light passing 

 through the slit is measured with a thermopile (or actinometer). In 

 the case of gaseous reactions the readings are taken when the cell is 

 empty and again when the cell is filled. In the case of solutions the 

 intensity of transmitted light is measured with a thermopile when the cell 

 is filled with a pure solvent and again when filled with a solution. The 

 difference in intensity in these two cases represents the energj^ absorbed 

 by the solute. It is convenient to have two reaction vessels of identical 

 dimensions placed side by side. One is empty or contains a solvent; the 

 other contains the reacting system. A slight correction is necessary for 

 reflections at the surface of a cell as given by Fresnel's well-known law 

 of reflection. With flat, plane surfaces and light striking at right angles, 

 the reflection loss at a gas-glass (or -quartz) interface is about 4 per cent 

 of the total intensity. The reflection loss at a liquid-glass interface may 

 be neglected. In precision work when the thermopile is placed back 

 of the reaction cell, some of this reflected light passes back into the reac- 

 tion chamber and requires a small correction. With three gas-glass 

 surfaces, one at the rear of the cell, one at the front of the thermopile 

 window, and a third at the back of the thermopile window, approxi- 

 mately 10 to 12 per cent of the light transmitted through the reaction 

 cell will be reflected back into the reacting mixture. 



The light which passes through the reaction chamber is measured by a 

 thermopile. Usually the surface of the thermopile receivers is about 

 1 mm. wide and about 1 cm. high and the area of the reaction vessel is 

 much larger. It is necessary then to converge, by means of a lens, all the 

 light which passes through the reaction cell so that it all falls on the 

 receivers of the thermopile. In case such a procedure is not convenient, 

 the thermopile must be moved back and forth and up and down across 

 the rear of the reaction chamber. In this way an integrated value for the 

 total light passing through the cell is obtained. It is comparatively 

 easy, as described on page 277, to make thermopiles having an area 

 1 cm. wide and 4 cm. high and a parallel beam of light issuing from a 

 monochromator or from a filter system can be of such dimensions as to 

 fall completely within the area of such a thermopile. When still larger 

 cells are used, it becomes necessary to integrate even with this large-area 

 thermopile. In accurate work it is usually necessary to thermostat the 

 reaction vessel and it is convenient to immerse the reaction chamber 

 and the thermopile in a vessel containing thermostated water. Copper 

 and brass are preferred for such a thermostat because any trace of iron 

 rust in the water produces a considerable change in the absorption of 

 blue and ultra-violet light. The thermopile is, of course, affected by 

 differences in temperature and it is advisable, although more troublesome, 



