Hygrometry . 401 
combined effects of its own property, and of the hygrofcopic 
law refpe cling evaporation which was known to me from other 
phenomena. 
79. As for the caufe of thofe ajiomalies which had been 
removed by the change of the procefs, the following phenome- 
non led me to difcover it. I obferved frequently, at times when 
my hygrometer , placed under the jar, flood at a considerable dis- 
tance from its point of extreme moifture , that a very fmali dimi- 
nution of heat was Sufficient to caufe, on the lower part of the 
vefl'el, the formation of a tarnijhed rim, extending one or two 
inches above the furface of the water , with a thin vanishing 
• edge. Having reflected on that phenomenon, from the me- 
chanifm I affign to the operation of fire in the very ad of 
evaporation I concluded, that in fuch a jlagnant air every eva- 
porating furface had an atmofiphere of extreme moifture, which 
extended as indicated by the tarnijhed rim ; and that it was 
only beyond that limit that reigned the other law, of a de- 
creafing moifture , correspondent to the increafing maxima of 
evaporation by an increafing heat . That new law of evapora- 
tion offered evidently an adequate Caufe for explaining the ano- 
malies obferved in the wet vefl'el ; for glafs retains concrete 
water very imperfedly, and it runs down very foon from 
many of the places which had retained it. Confequently, 
under fuch a partially wet vefl'el, and differently fo at different 
times, the inftruments muft be varioufly affeded by Scattered 
atmofpheres of extreme moifture . 
80. However, before I could truft that explanation, I 
wanted to Submit it to fome dired experiment ; and I Suc- 
ceeded by a means which at the fame time realized what M. 
de Saussure thought he had obtained, namely, to produce ex- 
treme moifture in 3 clofe Space, during any common temperature , 
Vol. LXXXI. G g g without 
