226 ATMOSPHERE IN RELATION TO HUMAN LIFE AND HEALTH. 



different forms which are only obscurely understood. There seems to 

 be a stage between the gaseous and the misty in which vapor is 

 condensed into very minute transparent motes or into a condition 

 corresponding to the critical state, the viscous interval, observed by 

 Andrews in carbon dioxide under great pressure. Just above this 

 critical point this gas behaved to some degree like its vapor and liquid 

 below it with regard to pressure. The behavior of water vapor under 

 varying pressure and when near saturation at different temperatures 

 would be an interesting though difficult subject for research. Dry 

 vapor is regarded by some experimental observers as diathermanous, 

 like air; yet we certainly find that what seems to be invisible trans- 

 parent vapor does largely arrest radiation from the earth. Therefore, 

 it would seem much of the vapor of the air, when near saturation, must 

 be in a condition bordering on mist or finely divided water. Only 

 beyond a certain size, maybe, or when dust is thick, do the particles 

 become large enough to give the effect of haze. It often happens that 

 a thermometer freely exposed to the sky on a fine night suddenly ceases 

 to fall, and rises several degrees without any apparent cloudiness or 

 diminution of the luster of the stars, but this rise, in the present writer's 

 experience, is a good indication of approaching rain after dry weather. 

 Whether the screen in the upper air which reflects the radiation from 

 the earth be a thin cloud or else vapor in a state of inchoate condensa- 

 tion, has not yet been ascertained. 



Haze, fogs, and clouds are caused by the tendency of vapor to con- 

 dense upon solid particles below a certain temperature. A change of 

 state from vapor to liquid or liquid to solid occurs much earlier in the 

 presence of "free surfaces" of other bodies than where these are absent. 

 Saturated air, as we call it, can hold no more vapor in ordinary condi- 

 tions, but apart from solids and dust particles it could contain much 

 more vapor without precipitation. Similarly, if water could be heated 

 by itself apart from solids and contained gases, it would rise high 

 above the boiling point without boiling, and would eventually explode; 

 so also the droplets of a cloud do not freeze, though many degrees 

 below the freezing point, until they touch a solid object. Dust in the 

 air offers the free surface which is required for condensation. Differ- 

 ent kinds of dust differ greatly in the power of compelling deposition. 

 Sulphur, magnesia, and common salt are, in the laboratory, at any rate, 

 powerful fog producers. In the open air sulphur seems to have little 

 appreciable effect; but salt, which is hygroscopic, or damp-attracting, 

 and pervades the atmosphere, plays an important part. Smoke, again, 

 or finely divided tarry matter, greatly favors fog formation, owing, 

 probably, to its strong radiative capacity and to its coating the water 

 globules so as to prevent evaporation. 



Suppose the motes of dust or salt in heterogeneous air to be radiat- 

 ing freely, and therefore to be colder than the air, and suppose each of 

 them to be frequently brought in contact with filaments of air and 



