Photometric Observations of the Sun and Ski/. 



275 



cides with C, it forms a rectangular trihedral of sensitised paper. 

 This is placed in a small exposure trihedral of cardboard, and covered 

 by a thin metallic trihedral in the trihedral angle of the octant. (I 

 make several of these trihedrals of sensitised paper, so as " in the 

 field " to take quickly a series of observations.) 



The trihedral of sensitised paper is, of course, carefully covered 

 up till the instrument is in adjustment; if then exposed to the action 

 of the sky for (say) 30 seconds, the readings on the quadrantal 

 planes MOS and IOS will be each 30 i a , and that on the quadrantal 

 plane MO I will be 30 . ^TT . <i a . 



36. 1 tried this octant actinometer on the 13th August, 1890 the 



*st day that the sky had been partially clear for a long time and 

 Iso (with a more imperfect sky) on the 15th and 16th August, 1890, 

 it Milverton, near Taunton. The exposures were all for 30 seconds. 



give the whole results. 



Table F. 



It is evident that these observations were interfered with greatly 

 haze or cloud; but it may be well to explain exactly how they 

 ?ere taken. 



A " sunstrip " was shaded first by the water-motion actinometer ; 

 le altitude of the Sun being known, the value of any line in this 

 in strip, in terms of the Dacca candle unit, was known by the aid of 

 Table B. 



The adjustment and working of the octant actinometer were found 



lot difficult. The readings in the two planes [V] and [W] were 



found practically equal ; the results are in the third column. These 



readings " were obtained directly by comparison with the " sun- 



