﻿Coloured Cloudy Condensation. 33 



proportion as the air is less dusty. Bearing this in mind, the 

 margin of the opaque zone at 20° for filtered air is not as much 

 above the atmospheric curve as one would anticipate. It 

 follows that the size of the particles producing coloured 

 cloudy condensation in atmospheric air is not necessarily 

 enormous when compared with molecular diameters — an 

 inference which I have already drawn* both from the 

 character of the colour phenomena and from the conditions of 

 condensation. 



15. Having obtained these preliminary results, I attacked 

 the subject on a much larger scale, using an apparatus very 

 similar to fig. 1 (with the phosphorus-tube removed), except 

 that the air, instead of being taken out of the atmosphere at c 

 and d, passed at those points through two large filters of the 

 type H (fig. 1). A large Root blower, actuated by a one- 

 horse-power gas-engine, forced the air through the system. 



Contrary to my expectations, this arrangement remained to 

 the last utterly iijefficieut. When the blower acted under low 

 pressures, the air entering the colour-tube was insufficient in 

 quantity. For higher blower-pressures and a more rapid 

 current of air f, the evidences of filtration were practically 

 absent. It was easy to trace the increase of dust-contents with 

 the velocity of the current of nominally filtered air, even 

 though high speeds were excluded by the nature of the case. 

 The nozzle of the original steam-jet being 0*16 cm. in diameter, 

 I replaced it by a finer one, 0*09 cm. in diameter, but without 

 advantage. It is noteworthy that for atmospheric air both 

 these jets gave identical results as to the location of the opaque 

 margin near the asymptote. The colours for the fine jet were 

 fainter, and together with the opaque field vanished at a lower 

 temperature, as one would suppose, seeing tha£ only one third 

 as much steam is available in one case as in the other. 



The best results for filtered air are, therefore, those of the 

 preceding paragraph. A sufficient degree of supersaturation 

 presupposed, in no case was there an absence of condensation ; 

 but as I cannot assert that the air used was rigorously pure, 

 it does not follow that I have reached the conditions f under 

 which the molecules themselves act as condensation nuclei. 



* Amer. Met. Journ. ix. pp. 507, 519 (1803). 



f Compare this with the similar experiences ot Mr. Aitken, in Trans. 

 Roy. Soc. Edinburgh, xxxv. p. 11 et seq. (1888). 



X The delicate question of purity comes into play in the other researches. 

 Thus condensation apparently without nuclei was produced by Aitken 

 (Trans. Roy. Soc. Edin. xxxv. p. 16, 1888), using the expansion method. 

 R. v. Helcuholtz failed to obtain it for exhaustions up to one-half atmo- 

 sphere (Wied. Ann. xxvii. p. 521, 1886), whereas Aitken's exhaustions 

 were only to three-fourths atmosphere (L c. p. 8). 



Phil Mag. S. 5. Vol. 38. No. 230. July 1894. D 



