C 42* 3 
being either too narrow or too ffiort. The propor- 
tion I had fettled between the capacity of this tube, 
and that of the ivory pipe, was deduced from the pre- 
liminary experiments I had made in the month of 
December (29); and had of courfe been found juft, 
as long as the fpring laded. But even before the 
new hygrometers were completed, the quickdlver had 
rifen in the fird fo as to run out of the top of the 
tube. This, joined to fome other previous obferva- 
tions, which had convinced me that the diminution 
of the humor is much more confiderable on moua- 
tains than in plains (76), induced me to fix the di- 
menfions of the tube of the hygrometer in the man- 
ner laid down in the defcription of the indrument. I 
had been in time to follow thefe dimenfions in the 
conftrudtion of my new hygrometers, fo that when 
they were brought from extreme humidity to the 
date of the air in my apartment in the month of Au- 
gud, the quickdlver did not rife too high in them j 
that is, it remained fufficiently below the top of the 
tube, to indicate leder degrees of humidity afterwards. 
60. The four new hygrometers have been con- 
druded with as little reference to each other, as if 
they had been made in different countries. By 
comparing them therefore, I have been enabled to 
judge of what might be expeded from the agree- 
ment of indruments of this kind. This is what I 
have found. 
When I have obferved them in places where it 
appeared likely that the humor would be equally dis- 
tributed among them, the utmod of their difference 
has been ufually from 19 to 21. Their greated 
height, for indance,, in my room with the windows 
