PAPER BY PROF. BEZOLD. 279 



Perhaps oue inigbt consider it a theoretical error that this behavior 

 was not from the very first made the object of investigation bat that 

 the mixture of such masses with other air was chosen as the starting 

 point of this study. But on the one hand this was the way by which 

 I was actually led to the whole subject, and on the other hand abbrevi- 

 ations and simplifications are hereby rendered possible that seem to me 

 sufficiently important to justify retaining this arrangement of the sub- 

 ject-matter. 



In order to study the behavior of such mixtures when left to them- 

 selves we have only to choose as a starting point the condition repre- 

 sented by the ordinate T 3 F 3 in Fig. 44, there considered as a state of 

 transition, and we arrive then, according to the same rules as above, 

 to the final condition T I and thus also to the final temperature T. 



Hence we recognize immediately " that mixtures of water and unsat- 

 urated air as soon as left to themselves must cool and so much the more the 

 further the vapor is from the point of saturation and the more liquid water 

 or ice is mixed with it." 



These considerations explain a phenomenon that I had frequently 

 observed, but concerning which I was until recently not certain 

 whether it might not be of a purely subjective nature. 



It had frequently happened to me en passing through strata of fog 

 such as fill the mountain valleys in the mornings of calm and subse- 

 quently clear days, that the impression of more severe cold occurs 

 precisely when we attain the upper limit of the fog as we ascend the 

 valley. Again it frequently occurred that just before the sun dissipates 

 the morning fog that collects in valleys or spreads over the lowlands, 

 the sensation of especial cold is experienced. 



Such impressions of the sensations can of course very easily lead to 

 error: but according to what has been above said, it is probable from 

 theoretical grounds also, that the temperature just below the upper 

 boundary of a dissolving layer of fog is lower than that of the layers 

 above an i below it. For when the sun begins to do its work at the 

 upper boundary of the fog there then occurs, first, relative dryness im- 

 mediately above this, and this relative dryness will, according to the 

 velocity with which evaporation of the fog particles takes place, partly 

 by diffusion, partly by direct radiation, also propagate itself to a 

 certain, although very moderate depth, in the layer of fog. 



But thereby (at least in many cases) the evaporation will be more ac- 

 celerated than would correspond to the increase of heat by direct radia- 

 tion, that is to say, the temperature must sink. 



This expectation, grounded partly on the impression of one's feeling, 

 partly on theoretical physical considerations, has now. during the writ- 

 ing of this memoir, received a confirmation by actual measurements. 

 For the communication of these measurements I have to thank Mr. 

 Bartsch von Sigsfeld, who, in a balloon constructed at his own expense, 

 has already undertaken many aerial voyages for scientific purposes 



