AND ITS CONGENERS. 213 



specks seen by us upon Hind Head 1 were the very, 

 kind needed for the production of the phenomenon. 

 But the observations of Ulloa place his white bow dis- 

 tinctly within the arc that would be occupied by the 

 ordinary rainbow — that is to say, in the region of super- 

 numeraries — and by the action of the supernumeraries 

 upon each other Ulloa's bow was accounted for by 

 Thomas Young. The smaller the drops, the broader 

 are the zones of the supernumerary bows, and Young 

 proved by calculation that when the drops have a 

 diameter of ^Vu^i or ToVo^ °f an inch, the bands 

 overlap each other, and produce white light by their 

 mixture. Unlike the geometric bow, the radius of the 

 white bow varies within certain limits, which M. Bra- 

 vais shows to be 33° 30' and 41° 46' respectively. In 

 the latter case the white bow is the ordinary bow 

 deprived of its colour by the superposition due to the 

 smallness of the drops. In all the other cases it is 

 produced by the action of the supernumeraries. 



The physical investigator desires not only to ob- 

 serve natural phenomena but to re-create them — to 

 bring them, that is, under the dominion of experiment. 

 From observation we learn what Nature is willing to 

 reveal. In experimenting we place her in the witness- 

 box, cross-examine her, and extract from her knowledge 

 in excess of that which would, or could, be sponta- 

 neously given. Accordingly, on my return from Swit- 

 zerland last October, I sought to reproduce in the 

 laboratory the effects observed among the mountains. 

 My first object, therefore, was to obtain artificially a 

 mixture of fog and drizzle like that observed from the 



1 Had our refuge in the Alps been built on the southern side of 

 the valley of the Rhone, so as to enable us to look with the sun 

 behind us into the valley and across it, we should, I think, have fre- 

 quently seen the white bow. 



