S. Cook — Spectrum of Aqueous Vapor, 26T 



Ir can only be explained as due to the very high vapor con- 

 cerned in the storm just ceasing. It is of course very nnnsnal 



for the air to clear bo Boon after rain ceases. It is to I)* 1 noted 



that the vapor line was not bo very much less intense on the 



and thin cloud than it was on the storm cloud an hour 



before when it was " raining furiously." Some three hours 



later, i. e.. at 3.15, the cloud had settled down again BO thick 



that spectroscopic readings became difficult, yet the line had 



diminished to an intensity of 1">. Whatever weight we may 



igm to those observations must be thrown in favor of the 



first cause mentioned, viz : that the great optical effect of 

 rin clouds is due to their being associated with a very high 

 altitude of vapor. A> 3,000 feet. of cloud seemed to exert no 

 great optical effect, the second cause suggested (multiple inter- 

 nal reflection) does not appear as the predominating source of 

 action. The result i> no more favorable to the theory of 

 abnormal action of saturated air. Much more conclusive 

 evidence could be secured if one of the observers were above 

 a Btorm cloud. A brisk shower would answer the purpose 

 quite well. With such conditions, the effect due to the cloud 

 would be isolated, and it would appear that the question would 

 be solved. Unfortunately, no such conditions presented them- 

 selves to US. 



It is rather remarkable that no amount of fog appears suffi- 

 cient to produce any spectroscope effect. At Hanover, where 

 most of the observations have been made, dense fogs are very 

 common, especially during the fall. Innumerable observations 

 have failed to detect more than a trace of action. Often in the 

 morning the darkness would be so great that one w T ould be very 

 uncertain as to the weather, the presence of clouds as well as 

 fog appearing probable. The sky might be so dark as to make 

 readings difficult, sometimes they were as difficult to make as 

 in the presence of a storm. Under such circumstances, the 

 spectroscope infallibly discriminated, rarely showing any line 

 at all in case of mere fog, while in the presence of storm cloud 

 it would be strong. The mere matter of brightness or dark- 

 ness of a stratum of cloud gives no clue as to the strength of 

 the line to be observed. Considerable attention has been given 

 to this point. As an example, [ will give the result of a com- 

 parison made Aug. 25 and 26, 1886, the observations being on 

 the summit of Moosilauke Mt. The 25th was stormy, rain 

 beginning to fall at noon, falling steadily till 2, and intermit- 

 tently afterwards. About 02 of an inch fell in all Observa- 

 tions taken on the cloud or rather in the cloud as the mountain 

 was enveloped constantly, during the lulls, i. e., when it was 

 not raining, gave readings of 16, 21, 24, 23, 25 and 20. On 

 the next day no rain fell but otherwise the conditions of things 



