AND ITS CONGENERS. 215 



away, leaving behind it a host of minute liquid 

 spherules floating in the beam. An intensely-coloured 

 circular rainbow was instantly seen in the air in front 

 of the observer. The primary bow was duly attended 

 by its secondary, with the colours, as usual, reversed. 

 The opening of the valve for a single second caused 

 the bows to flash forth. Thus, twenty times in succes- 

 sion, puffs could be allowed to issue from the boiler, 

 every puff being followed by the appearance of this 

 splendid meteor. The bows produced by single puffs 

 are evanescent, because the little globules rapidly dis- 

 appear. Greater permanence is secured when the valve 

 is left open for an interval sufficient to discharge a 

 copious amount of drizzle into the air. 1 



Many other appliances for producing a fine rain 

 have been tried, but a reference to two of them will 

 suffice. The rose of a watering-pot naturally suggests 

 a means of producing a shower ; and on the principle 

 of the rose I had some spray-producers constructed. 

 In each case the outer surface was convex, the thin 

 convex metal plate being pierced by orifices too small 

 to be seen by the naked eye. Small as they are, fillets 

 of very sensible magnitude issue from the orifices, 

 but at some distance below the orifices the fillets shake 

 themselves asunder and form a fine rain. The small 



1 It is perhaps worth noting here, that when the camera'and lens 

 are used, the beam which sends its ' effective rays ' to the eye may 

 not be more than a foot in width, while the circular bow engendered 

 by these rays may be, to all appearance, fifteen or twenty feet in 

 diameter. In such a beam, indeed, the drops which produce the bow 

 must be very near the eye, for rays from the more distant drops 

 would not attain the required angle. The apparent distance of the 

 circular bow is often great in comparison with that of the originat- 

 ing drops. Both distance and diameter may be made to undergo 

 variations. In the rainbow we do not see a localised object, but 

 receive a luminous impression, which is often transferred to a por- 

 tion of tbe field of view far removed from the bow's origin, 



