278 Meteorological Observations at Hudson, Ohio. 
before each observation. Very little confidence can be placed in 
the observations of this instrament made by an unskillful person ; 
but with a judicious observer, I think this instrument is as much 
to be relied upon as any hygrometer with which I am acquainted. 
The second objection is well founded, but is not very serious. 
It is rare that the observation requires more than three or four 
minutes, and the same objection lies against nearly every direct 
method of determining the dew point. When observations are 
to be made only a few times in a day, the loss of time is not 
great ; but where hourly observations are required, I should pre- 
fer Prof. Bache’s hygrometer to any thing else [I have yet seen. 
The observation with the wet-bulb hygrometer is very expedi- 
tiously made; but this furnishes the dew point only by computa- 
tion, and the computation is as laborious as the observation with 
Daniell’s hygrometer. 
The third objection may be true for some climates, but I 
doubt if it be so for any part of the United States. Daniell’s 
hygrometer has been observed at Hudson for seven years, every 
ay at 3 p.m., and the experiment has never failed. Twelve 
times the dew point has been more than 30° below the temperature 
of the air, and once the difference amounted to 36°. It is doubt- 
ful whether the difference would ever be found much greater for 
any part of the United States. In the hands of many observers 
the experiment would fail at such times. It can only succeed 
with the aid of good ether and dextrous management. Most of 
the ether of commerce is unfit for the purpose, and can be used 
advantageously only after distillation or washing. With the best 
ether I could command, I have sometimes found all the contents 
of the lower ball distilled over into the upper, without the depo- 
sition of dew. In such a case, I immediately invert the instru- 
ment, and drive back the ether into the lower ball, and repeat 
the operation before the thermometer has had much time to rise. 
Without this precaution the experiment would sometimes have 
failed. The wet-bulb hygrometer has a seeming advantage ovet 
Daniell’s in this respect that you can easily get an observation. 
This is a method well deserving attention, yet its theory can 
hardly be considered as sufficiently settled to entitle it to the 
same confidence as a direct method of getting the dew point. 
In the observation with Daniell’s hygrometer I have often 
noticed the following curious fact. As ether is applied, the ther- 
