THE STUDY OP TROPICAL SUNLIGHT. 3 



and Kozak," in a ranjj^e of temperature observations extending from 4'' to 

 80°, using uranyl nitrate and oxalic acid, could observe no acceleration 

 with increasing temperature, and observe "this is probably the first 

 photochemical reaction with such a very small temperature coefficient, if, 

 indeed, the latter is not equal to naught." The mixture, heated in a 

 water-bath in the dark, shows no decomposition, and flasks containing 

 uranyl acetate-oxalic acid solutions, when covered with thin silver foil and 

 exposed to the full effect of the sunlight, likewise show no change even 

 though they become markedly heated. 



Therefore, it seemed safe to assume that a temperature coefficient would 

 be negligible or nonexistent in comparative measurements, but we were 

 once more brought back to this phase of the reaction by an observation 

 made by Mr. M. Barrowcliff, of the Institute for Medical Research at 

 Kuala Lumpur,* who, in carrying on measurements with the standard 

 solution, called our attention to the fact that, in his opinion, a temperature 

 coefficient in reality existed. 



The entire matter was open for reinvestigation, and Mr. W. C. 

 Holmes, of the laboratory of organic chemistry of the Bureau of 

 Science, conducted a careful series of experiments confirming 

 Mr. Barrowcliff's results. There is very little difference between 

 the reaction at 30° and at the boiling point, as Bacon had shown, 

 but between 2° and 30° there is a marked increase in the rate 

 of decomposition with the rise of temperature, so much so, that 

 decomposition at 2° is only 60 per cent of that at 30°, whereas 

 that at 30° is 95 per cent of the decomposition at 75° ; above this 

 point, there is even a diminution of the rate with increased 

 temperature. Between ordinary points of measurement in our 

 climate, say from 25° to 35°, the temperature coefficient may, 

 therefore, be neglected; where measurements at lower tem- 

 peratures are made, the correction would need to be applied, 

 although in comparing tropical climates with those of tem- 

 perate zones this temperature coefficient would work in the 

 direction of greater contrast and apparently lesser insolation 

 in the latter, so that, if such contrast were not evident on com- 

 paring measurements, the coefficient could be neglected in draw- 

 ing comparative conclusions. Wherein the difference between 

 these results and those obtained by Bruner and Kozak lies, 

 we have not yet determined.^ 



Other factors influencing the reaction must also be considered 

 before the method can be used in a series of comparative meas- 

 urements ; these are : The nature of the background and the size 



• Loc. cit., 357. 



' Communicated by letter. 



' Possibly, in the dilute solutions we use, the temperature coefficient of 

 the decomposition of oxalic acid alone becomes evident. It scarcely seems 

 practical to use greater concentrations of uranyl acetate in cold climates. 



