THE STUDY OF TROPICAL SUNLIGHT. 7 



SO slight that we can say that practically the photocatalytic action 

 in all is the same. 



Unfortunately, only a few data have reached us from Tucson, 

 Arizona (32° 12' north) (Table IV), and these for the month 

 of October. They show a maximum of 13.4, or 4.4 less than 

 that of Manila, and a minimum on one day of 7.7 or 2.5 greater 

 than the average at this place. The temperature during the 

 observation averaged about as it does here (28°.7), and higher 

 than at Honolulu. Doubtless, when a longer series of obser- 

 vations is at hand from this interesting point, we will discover 

 many days in Tucson where the maximum is as high as, or higher 

 than, it is here, and an average about the same. 



The data from Washington (Tables V and VI) need a little 

 more careful analysis, as the methods followed were not always 

 identical with the ones adopted by us as a standard, the hours 

 of insolation were not always the same and recalculations need 

 to be made in that respect. Nevertheless, so far as they are 

 comparable, the results show that Washington, which has a 

 winter climate, presumably more atmospheric disturbances, and 

 many cloudy days and possibly but few absolutely clear days, can 

 show at times as much effect as the four places discussed and 

 an astonishingly high average of 11.80. One day in September 

 gave an hourly decomposition, between 8.45 and 12.15 in the 

 morning, of 19.14 per cent, and making allowance for the greater 

 concentration of uranyl acetate used by Doctor Bacon, it would 

 still be close to 16.00 ; and, in February, between the hours of 9 

 and 1, a decomposition, which, making allowance for the temper- 

 ature coefficient and for the concentration of uranyl acetate, 

 would certainly in Washington show but little difference between 

 summer months (July, August, September) , 11.0, and the winter 

 months (December, January, February), 10.0. Making allow- 

 ance for the greater concentration of uranyl acetate used by 

 Bacon, the totals in Washington are lower by about 33 per 

 cent than in Manila, excepting the one month, November, in 

 Manila with an average of 9.97. 



The results in Khartoum, Sudan (Table VII), are extremely 

 interesting and, perhaps, the most instructive of the series. 

 Khartoum is close to the desert and in about the same latitude 

 as Manila. We find here, in observations extending through the 

 months of September, October, and November, an average of 

 17.6 as measured by a standard quartz flask, or as much as 

 5.15 higher than Manila and 2.3 higher than Kuala Lumpur, 



