216 THE RAINBOW 



orifices are very liable to get clogged by the particles 

 suspended in London water. In experiments with the 

 rose, filtered water was therefore resorted to. A large 

 vessel was mounted on the roof of the Royal Institu- 

 tion, from the bottom of which descended vertically 

 a piece of compo-tubing, an inch in diameter and 

 about twenty feet long. By means of proper screw- 

 fittings, a single rose, or when it is desired to 

 increase the magnitude or density of the shower, a 

 group of two, three, or four roses, was attached to 

 the end of the compo-tube. From these, on the turn- 

 ing on of a cock, the rain fell. The circular bows 

 produced by such rain are far richer in colour than 

 those produced by the smaller globules of the con- 

 densed steam. To see the effect in all its beauty and 

 completeness, it is necessary to stand well within the 

 shower, not outside of it. A waterproof coat and cap 

 are therefore needed, to which a pair of goloshes may 

 be added with advantage. A person standing outside 

 the beam may see bits of both primary and secondary 

 bows in the places fixed by their respective angles; but 

 the colours are washy and unimpressive. Within the 

 shower, with the shadow of the head occupying its 

 proper position on the screen, the brilliancy of the 

 effect is extraordinary. The primary clothes itself in 

 the richest tints, Avhile the secondary, though less 

 vivid, shows its colours in surprising strength and 

 purity. 



But the primary bow is accompanied by appearances 

 calculated to attract and rivet attention almost more 

 than the bow itself. I have already mentioned the 

 existence of effective rays over and above those which 

 go to form the geometric bow. They fall within the 

 primary, and, to use the words of Thomas Young t 

 "would exhibit a continued diffusion of fainter light], 



