In the course of his experiments, Dr Wells observed, that water exposed early in the evening in the open air to the sky, lost a little weight during a clear night. This he imputed to evaporation taking place before the water had been cooled enough to condense the vapour of the atmosphere, and to the weight gained afterwards being insufficient to compensate the previous loss. He ex posed, therefore, water to the influence of the sky, until it was cooled to 34°; of this he put two ounces into each of two china saucers, which had also been exposed to the air, and then placed the saucers upon the straw-bed. In the morning a thin cake of ice was formed in both sau cers, one of which had gained 23, and the other 3 grains in weight. Dew was also copious in the night. At one time grass was 93°, and the exposed part of the straw bed 12°, colder than the air.
Dr Wells regards the result of these experiments as a decided proof, that the cause of the formation of ice, in such circumstances, cannot be evaporation ; but be fore coming to this conclusion, he ought to have shewn, that, during the experiments, there was DO evaporation from the surface of the ground, as well as from the wa ter in the pans. For it is easy to conceive, that the evaporation from the extensive surface of the fibres of the grass, might be sufficiently great to cool the air in contact with them below the freezing point, and that this tcperature might afterwards be communicated to the air immediately above the pans, either by conduc tion, or the actual transference of the cooled particles of air. Taking this view of the matter, we can explain the reason why ice was formed in the pans, though the water which they contained suffered no loss by evapo ration, or even gained a small addition to its weight, as part of the moisture which evaporated from the surface of the ground might have been condensed over the pans. Dr Wells mentions, on the authority of Sir R. Barker and Mr Williams, both of whom witnessed the process in India, that, for the complete success of the experiment, it is necessary the air should be very still ; and he adds, that " Ind, which so greatly promotes evaporation, pre vents the freezing altogether," without seeming to be aware, that this fact is still more irreconcileablc with his own hypothesis, it being well known, that radiation is not affected by a transverse current of air. But, ad mitting the opinion, that evaporation is the cause of the reduction of temperature, a good reason may be assign ed why an agitated state of the atmosphere, though it tends to increase the evaporation, may be unfavourable to the formation of ice ; because the air in contact with the surface, and whose temperature is reduced by the evaporation, is in that case carried off, and mixed with the general mass, before it has time to communicate its temperature to the water in the pans. No reason can
be given, on the theory of radiation, why the eonge lation succeeds best when the pans are placed in exca vations ; but, on the theory of evaporation, these exca tions afford receptacles for the cooled air, which, ha ving its temperature diminished, is increased in specific gravity, and thus settles in the lowest situations. The slightest agitation of the air would prevent this. Though we state these objections against the hypothesis of Dr Wells, we are not disposed to say, that radiation has no share in the effect, or, that the theory which ascribes the reduction of temperature entirely to evaporation, is al together free from difficulties. Several experiments, performed with the utmost attention to accuracy, are still necessary for the full elucidation of the subject.
In some of the arts, particularly dyeing and calico printing, it is necessary to produce a quick evaporation without exposure to the light of day. This is sometimes done by suspending the wet stuffs in buildings for the purpose, having a great number of perforations, so as to exclude the light, hut allow a continual current of fresh air to circulate through them ; at other times, the same thing is effected by exposing the stuffs in a stove, which has usually a fire, and long flues for diffusing the heat through the room. The stove is furnished with several .openings at the top, to allow the escape of the air after it becomes charged with vapour.
The desiccation of other bodies must be carried on, as much as possible, without the agency of heat : Thus the form, as well as the colour of delicate plants, upon which so much of their value in the preserved state de pends, would be in a great measure destroyed by expo sing them to the heat of a stove. It has therefore been proposed, to dry them in an exhausted receiver, by sus pending them over a vessel containing sulphuric acid, the muriate of lime, or any other substance which has a strong attraction for humidity, and allowing them to re main in that situation, until the whole of their moisture was evaporated. it has even been proposed, though with less probability of success, to dry gunpowder in the same manner. On the subject of evaporation, see Phil. Trans. Nos. 192, 407. vol. lv. p. 182; vol. lxvii. p. 134. and 257. D. Luc, Idees sur la Meteorologic. Phil. Trans. for 1792. D ilton's Essays in the Manchester Memoirs, vol. v. Saussure's Essays on Hygrometry. Murray's Chemistry, p. 705. Dr Wells on Dew. (A)