200 



Mr. W. 



10. The observations of Table (A), and numerous other similar 

 observations, were taken with great care, the strips being read i In- 

 same evening. The strips taken on separate days, were also compared 

 with each other; it was thus found that the numerical values for flu- 

 chemical action were the same, with different paper, and with different 

 candles. In England, I have, within the last two years, made similar 

 observations to those at Dacca twenty-five years ago, and I submit 

 three of the strips taken ; these have been fixed, and have conse- 

 quently changed both in density and in colour, and are submit 

 merely for explanation. The photographic sensitised paper, now pre- 

 pared in England, keeps in the dark for months unchanged, and 

 renders constant reference to the candle standard unnecessary. But 

 by the aid of the table (B) which immediately follows, I could always 

 in 1889 and 1890 recover the standard unit, by an observation of the 

 Sun better than from the candle. 



11. The chemical action of the Sun alone, is got in a dark room, 

 arranging a vertical slit, so that the Sun's light falls exactly down the 

 strip of paper, which I expose at right angles to his rays. To get the 

 chemical action of the Sun and sky (i.e., the portion of the visible 

 sky within 90 of the Sun) together, the exposure is completely in 

 the open. The chemical action of the sky (i.e., the resultant actioi 

 on the plane at right angles to the Sun of that portion of the visibl 

 sky within 90 of the Sun) is got by an exposure in the open, a 

 vertical stick having been arranged so that its shadow should just 

 cover the exposed strip. 



As I took each of these three kinds of observations, giving mnnr 

 rical results , /3, 7 respectively, I was enabled from the simpl 

 formula -|-<Y = /3 to check my observations, to test the closeness with 

 which the strips could be read certainly, and to show again that an 

 intensity of 4 acting for 1 second has the same effect as the intensity 

 of 1 acting for 4 seconds. 



12. I found, as Roscoe, working in a less pure atmosphere, fom 

 in a still greater degree, that observations very close to the horii 

 were not to be depended upon. Also, in the cold weather at Ds 



at which season alone the sky was sufficiently clear, the Sun did nc 

 attain a greater altitude than about 45. The flat roof of my hous 

 offered nearly a complete hemisphere of unclouded blue ; neverthe- 

 less, I know that the full effect of the band of sky near the horizon, 

 must have been to some extent interfered with by haze ; and the 

 constants in some of the tables that follow will be, in a small degree, 

 affected by this cause. 



13. The following Table (B) is shortened, from one which I drew 

 up and printed photographically at Dacca in 1865. It represents tlie 

 mean result of very numerous observations, taken at altitudes of the 

 Sun between 5 and 45. From 45 to 90, the table has been 



