10 ATMOSPHERIC ACTINOMETRY 



these solutions can becunie sensitive in a ooncentrateil state ami [nvserve this sensi- 

 tiveness even aftei- being diluted. One can tiien provide a mother-liquor, so to 

 speak, whirli nia\ be made sensitive and which afterwards may be diluted as 

 necessity arises. Ordinarily I used to pi'epare a UDiiual sdliitiDii to the amount of 

 several litres, containing 13 grammes per litre, wliicli I kept for some weeks under 

 diffused light, and subsequently diluted, in fractions, to the twentieth or fortieth 

 degree. One litre of this mother-liipiid, rendered duly sensitive, may thus serve 

 for 2000 tests. In all the comparative experiments which will be mentioned in 

 this memoir, I have always taken pains to work with identical liquids and such as 

 had the same sensitiveness. 



We are now possessed of our actual working process, which amounts to 

 this : To expose to the sun during the day a shallow dish, containing 20 c. c. of a 

 half-deci-uorraal solution of oxalic acid, which has become sensitive by time, and to 

 measure at the close of the day, by a titration with lime-water, the quantity of acid 

 which has disappeared by oxidation. 



Let us now see what results have been obtained by this process. 



ACTINOMETIIIC MEAStJREJrENTS. 



Since the year 1885 I have made seveial series of actiuometric measurements, 

 especially during fine weather and at times when I was sufficiently master of my 

 own time to secure to them the regularity which they require. All these experi- 

 ments, made at different times and at ditt"ei'ent places, are not absolutely alike, since 

 the solutions used might have undergone some change. But such variations amount 

 to little from one year to another, and to almost nothing in the course of the 

 same year, as I have been able to determine repeatedly ; for every time when I 

 changed the solution, I exposed simultaneously two or more vessels with old and 

 with new material, and I always found that the solar combustion was the same for 

 both, up to that degree of approximation which the measurements demand. 



While operating with two or more vessels containing one and the same liquid, 

 it does not always happen that we tiii<l the same result for all at the close of the 

 day. There are irregularities in the [)rocess, some of which will be explained 

 presently, while the others have until now defied all efforts at explanation, so sud- 

 den are they and so exceptional. There is no other remedy for this than to elimi- 

 nate such out-of-the-way cases, which are always rare, making every day a ti'ial with 

 3 or 4 vessels and keeping, of the figures thus obtained, only those which are con- 

 cordant. 



It is in this way that the following observations have been made. For each of 

 them a record hjis been kept of the proportion of oxalic acid burned in 10 c. c, of a 



