2,6 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 65 



higher records. Since then the apparatus has been wholly rebuilt, with Richard 

 clocks, and the best possible driving mechanism, so that backlash of the drum 

 is nearly eliminated. 



Neither you nor I have read, or ever can read, the pyrheliometer outside the 

 atmosphere. It is now proposed to cause automatic pyrheliometers to observe 

 as high up as possible. In the interest of learning the truth I beg that you will 

 be so good as to suggest to me wherein the proposed experiments are likely to 

 fail, so that all possible precautions may be taken against failure. Undoubtedly 

 it will be impossible to get results to 1 per cent, but — 



6. Do you see any reason why the experiments should not be decisive as 

 between a solar constant of 1.9 — 2.0 calories and one of 3.0 — 4.0 calories? 



I await with much interest your replies to my six (6) questions, and any 

 suggestions you may have the goodness to offer. 



In response to this communication, Mr. Very was kind enough to 

 send two letters which contain very valuable suggestions. We quote 

 a portion of the letters as received. 



(a) Without actually experimenting myself with such actinometric appa- 

 ratus as you use, I should not care to express an opinion as to its efficiency. 



(b) I regard the upper isothermal layer of the atmosphere as due mainly to 

 local heating through absorption of solar radiation. Until we get above that 

 layer, I should expect to find increment of solar radiation with each increase 

 of altitude. It seems to me improbable that this limit will be reached at 

 40 kilometers. 



(c) Any plan for a high level measurement of solar radiation which has 

 even a small prospect of success may be worth trying. It is to be regretted 

 that yours involves the local application of electric heating, which seems to me 

 very risky and liable to produce all sorts of complications and unforeseen 

 results. ... I would suggest that ascension should be made at night with a 

 little electric lamp to give the record, to see what sort of a record you would 

 get when the sun is away. The combination of night and day records 

 might enable you to eliminate some errors inevitable in the method. ... If 

 your disk and its attachments are too massive four minutes exposure may not 

 be long enough. You cannot use a very long exposure because the balloon 

 ascension ends too soon. It behooves you therefore to have your thermometer 

 and disk made on the smallest possible scale. Another thing which may be 

 unavoidable in your construction is the very circumscribed protecting case. 

 The same instrument may read differently in a wide, roomy case. . . . The 

 knowledge of how such an apparatus as you are proposing will behave in the 

 absence of the sun seems to me almost indispensable. Thus I should be 

 apprehensive that the interpositions of the metal cone above the heat-measuring 

 disk will act as a wind shield to some extent. There will, therefore, be less 

 cooling from contact with the air during shade that there would be if the wind 

 effect were constant, and the fall of temperature in shade will be too small in 

 the day observation. At night there might even be a rise of temperature when 

 the cone is interposed, and it is desirable to learn whether this is so, and the 

 amount of the change. . . . During the most rapid part of the ascent, the instru- 

 ment is exposed to a strong resultant air current, which may exceed 7 meters 

 per second. This powerful wind blowing directly upon the face of the 



