METEOROLOGY. 370 



found a definite very small quantity of dust M'hich is especially favora- 

 ble for the development of color by the process of diffraction, and this 

 can therefore be specified as the dust favorable to optical phenomena. 



2. The absolute maximum of color due to diffraction is in moist air 

 conditional upon the simultaneous occurrence of this optical dust with 

 the maximum relative humidity of the air stratum in which it is con- 

 tained. 



3. The twilight colors depend principally upon diffraction in moist 

 air. (Z). Jf. ^., I, p.34.) 



384. Professor Kiessling, of Hamburg-, has collected together the ob. 

 servations of others and compared them with his own observations on 

 the influence of artificial fog on direct snidight, and the consequent ex- 

 planation of the twilight i)henomena. lie uses a glass vessel with sev- 

 eral apertures in connection with a Sprengel or water air pump ; within 

 this vessel he forms fogs of varying degrees of intensity, and examines 

 the dillVaction phenonjeiia when a beam of sunlight falls upon the fog 

 from the heliostat. besides the fog fj om aqueous vapor, he also utilizes 

 the sulphate of ammonia, phosi)b()ric acid, gunpowder, and other sub- 

 stances i)rodncing very dens-^ smoke oi- fog ; lor instance, with sulphate 

 of ammonia a very dense smoke is ])rodaced consisting of exceedingly 

 small hard particles of this salt; after twenty or thirty minutes of set- 

 tling a thick white dust covers the base and sides of the vessel. The 

 sun, seen through the cloud, does not blind the eye, but shows a re- 

 markable change of colors. At the first moment of its formation the 

 sun appears of a brilliant dark copper color, but changes its color 

 rapidly, first to a violet tint and then through dark carmine red into a 

 brilliant azure blue. This change of colors occupies in dry air about 

 two minutes, but in moist air scarcely twenty seconds; similar phenom- 

 ena take place with other heavy vapors. If the lighL is allowed to fall 

 upon a white screen it is seen to be at first of a copper-brown, which 

 afterward becomes blue. If in place of these chemical dust-clouds we 

 allow a stream of aqueous vapor to escape from a steam boiler, the cloud 

 of fog shows an intense blue as soon as the above beam of blue solar 

 light falls ui)on it ; if the stream of vapor is, however, properly man- 

 aged, it can be made to exhibit in itself all the changes from brownish 

 red to blue, which, however, very rapidly disappear on account of the 

 dissipation of the cloud. In order to accomplish this best the stream of 

 steam must be horizontal and mixed with a feeble current of very cold 

 air drawn over ice. It has not yet been possible to Kiessliug to experi- 

 mentally reproduce the green colors of the sun. 



If the vessel within which the diifraction ])heuomena are produced is 

 full of moist, perfectly filtered air, entirely free from dust, then a fog can 

 be produced by simply lowering the temperature, which is, however, 

 only visible in direct sunlight. This fog consists of very small, scat- 

 tered, rapidly moving particles of vapor, which, however, in transmitted 

 light show not the slightest trace of colored diffraction rings, evidently 



