C 429 J 
fhut, has hitherto been 94, 994., roof, 105I, in the 
fame moment; which is pretty nearly in the pro- 
portion of 19 to 2 1, between the hygrometer which 
remains at the loweft, and that which is at the 
higheft. 
61. Befides this difference between the relative al- 
titudes of thefe four inftruments, I have obferved 
another kind of irregularity in them, which is, that 
they do not always preferve the fame proportion to 
each other. Thefe variations are undoubtedly in 
part owing to the caufe itfelf of their motions; that 
is, to the unequal diftribution of humor even in 
places very near each other ; but I have reafon to 
afcribe part of them to fome defed in the inftru- 
ments themfelves. I fhall hereafter return to thefe 
caufes, and give them a clofer examination. 
Considerations on the Degree of Accuracy* 
that has been obferved \ 
62. Notwithftanding the defeds I have mentioned 
were rather evident, I was not diffatisfied with this 
firft trial. I never imagined that I had forefeen every 
thing, and confequently could not exped to arrive 
at a fufhcient degree of exadnefs without the help 
of experience ; the irregularities therefore which ap- 
peared in the execution, did not make me defpair of 
being able to perfed that inftrument. 
63. My hopes in regard to this were at firft only 
grounded upon general refledions. I recolleded 
what the barometer and thermometer had been when 
they firft came out of the hands of their inventors; 
and obferved that in fome refpeds they were more 
Kkk 2 irregular 
