1882.] On Actinometrical Observations made in India. 53 



20. Next, I may first mention that by the " black bulb " thermometer 

 is intended one of those unsatisfactory instruments, of which the bulb 

 is blackened and enclosed in a fixed glass sphere some 3 inches in 

 diameter, and that the same identical instrument was exposed and 

 read on the same spot in 1879 to 1881. ISTow, according to what has 

 been said, the actinometrical summer result, if true, indicates a 

 sensible decrease of radiation compared with autumn, but the black 

 bulb shows exactly the reverse, and reads 19° higher in the summer ! 

 There is also a similar contradiction between 1879 and 1880, where 

 the black bulb gives a rise of 13 to an actinometrical fall. I confess 

 my inability to account for these contradictions, which, however, I 

 cannot regard as of serious importance, unless it be affirmed that 

 black bulb thermometers are reliable : in this case there arises the 

 question why the actinometer and the black bulb should be at such 

 utter variance ? 



21. In concluding this paper, I venture to make a few remarks in 

 respect to actinometrical observations. It is astonishing how con- 

 siderably one's notion of the number of days suitable for actinometrical 

 work undergoes modification after careful watching. I speak, of 

 course, only of visible vitiating causes ; of the invisible I know 

 nought, if ought be known. And the presence of the former causes 

 I now believe may be detected by a careful observer, at any place 

 I am acquainted with, on several days in the year : no doubt one 

 locality is far more suitable than another, but at the best a number of 

 days will occur annually, when observations if made are worse than 

 useless, for they are misleading. Hence in any project for continuous 

 actinometrical work, the stations should be selected at least in pairs, 

 so that the omissions at one of the two may be supplemented at the 

 other. 



22. Another point is this : suppose that at a given station the sun's 

 meridional zenith distance at lowest is £, and at highest is Z, then the 

 system of observation I suggest is as follows : — 



(1.) Observe daily from 15 minutes before to 15 minutes after 

 apparent local noon. 



(2.) Observe daily a series of 6 sun and 5 shade observations when 

 the sun on each side of the meridian is at zenith distance =g for the 

 mean of the set. 



(3.) Observe a series of 6 sun and 5 shade at every half hour from 

 9 a.m. to 3 p.m., if possible, on every day corresponding to a change 

 of 3° declination between g and Z ; thus if the sun's meridional zenith 

 ranged in course of the year between zenith distances 7° and 54° 

 (about), then by preference let the days of observation for long series 

 be those days when the meridional zenith distances are successively 

 7°, 10°, and 13° ... . 54° (about). 



23. By means of (1), (2), (3) a complete connexion will be estab- 



