432 On the Deficiency of Rain in an elevated Rain-yauge. 



supernumerary bows in the lower part, but only in the upper, 

 Arago argues that the condition of efficacious interference of the 

 drops must have been destroyed in descending into the lower 

 part of the atmosphere. " Therefore," he concludes, " the drops 

 of rain must have much increased in size*." Obviously this does 

 not in the least follow ; for the condition of efficacious interference 

 is uniformity of sizef ; and uniform drops, condensing moisture 

 upon themselves, or evaporating in the same circumstances, will 

 remain uniform in size. The disappearance of the supernume- 

 rary bows near the surface no doubt arises from the more dis- 

 turbed current of air there causing the drops to encounter each 

 other and coalesce irregularly, so that some drops are produced 

 two or three times as large as the others. 



26. Distant showers of rain are often seen distinctly to evapo- 

 rate, and sometimes entirely vanish during their fall; but I have 

 never observed or heard of a shower being observed to increase 

 in density visibly during its descent. 



27. It is now only right to add that both Arago J and Prof. 

 Phillips have recorded unequivocally that a deficiency of rain in 

 the upper gauge occurs even during a perfect calm. We have 

 already quoted one such observation by Prof. Phillips § ; and two 

 others are found in his second paper on this subject ||. Prof. 

 Phillips, indeed, considers that falling rain itself produces a 

 downward current of air, which, it is just conceivable, might, by 

 flowing over the sides of the upper rain-gauge or its support, 

 deflect the rain. Again, while a perfect calm prevails on the 

 ground, a gentle wind is usually blowing at the top of a lofty 

 tower. As my explanation of the deficiency of rain in an elevated 

 gauge is certainly inapplicable in a calm, I confess that my 

 hearers must choose for themselves between considering two dis- 

 tinguished scientific observers capable of mistake in the observa- 

 tion of wind and calm on the one hand, and overturning some 

 of the best established facts of physical science on the other hand. 



28. If the present explanation be accepted, all observations by 

 rain-gauges elevated or exposed to wind must be rejected as 

 fallacious and worse than useless. But it is improbable that the 

 error in a gauge with its mouth not more than one or two feet 

 above the ground is worth considering. Still I believe that 

 during a heavy shower almost all gauges lose a little rain by 

 splashing, and it is worthy of consideration whether more accu- 



* Annuaire du Bureau des Longitudes, pour l'an 1836, p. 300. 



t Herschel's 'Meteorology,' p. 219. It seems likely, however, that 

 Arago argued upon some other view of the cause of this phenomenon, 

 which has been much misunderstood. 



X Annuaire du Bureau des Longitudes, pour l'an 1824, p. 160. 



§ Sec above, paragraph 13. 



11 Report of the British Association, 1834, Trans. Sections, p. 561. 



