﻿Coloured Cloudy Condensation. 31 



the right of ABC is coloured, merging into colourless ; the 

 field to the left of ABO is opaque. 



As the air becomes more and more dust-laden, the yellow 

 territory encroaches on the blue, so that for unusually dusty 

 atmospheric air the pair of curves ABC has changed into ADC. 



In the same way the yellows will continue to advance upon 

 the blues for each successive (now artificial) increment of the 

 dust-contents of the air, until eventually the blues have been 

 quite crowded out of the field, and the whole territory is per- 

 sistently yellow at all temperatures and pressures. In other 

 words, the asymptote of the curve descends with increasing 

 dust-contents, while at the same time the curve BC moves 

 bodily to the right, so that BC finally coincides with the 

 co-ordinate axes of pressure and temperature. This at least 

 is the essential feature of the phenomenon so far as I now 

 understand it. Subsidiary details will be brought forward at 

 some other time. 



It is to be remembered that the particular curves, figs. 2 et 

 seq., apply primarily to the particular jet discharging into the 

 given tubes. Nor can any attempt as yet be made to graduate 

 the apparatus ; for a comparison with the Aitken dust-counter 

 is not legitimate, unless it can be proven that any given 

 class of particles * occurs proportionally to the total number, 

 certainly a hazardous hypothesis. For the present the height 

 of the curve at points above 28° (asymptote), expressed in 

 centimetres of mercury (pressure) , is the empiric dust-indicator. 



14. Filtered Air. — A grave difficulty is encountered in 

 using filtered air, inasmuch as the supply cannot be obtained 

 in sufficient quantity without employing very cumbersome 

 apparatus. I have helped myself provisionally by using a tin 

 tube G (fig. 1) with the end n n large enough to fit snugly 

 into the air-hole C when the heater is removed. G is filled 

 with cotton throughout about 20 cm. of its length in the 

 usual way. The end nn is closed with a sieve of brass wire 

 gauze, while the end m is closed with a perforated rubber 

 cork, through which an influx tubulure, o, projects. This is 

 connected with an ordinary gas-tank, containing highly com- 

 pressed air. 



When the tube'C (fig. 1) is closed with the filter G (no 

 gas passing through it), the field of A A at once becomes 

 clear ; at the same time, however, its temperature rises to the 



* Cf. §14. Evidence is also adduced in my earlier paper (Am. Met. 

 Jouvn. March 1893, §19) that it is only the smallest of dust particles — 

 i. e. particles whose dimensions are limited either way and of the order 

 of a few hundred molecular diameters — which induce coloured cloudy 

 condensation. 



