354 NOTES OF A NATURALIST. 



mercurial barometer is not indicated on the face of 

 the instrument. Assuming that the aneroid is com- 

 pensated for variations of temperature, and I have 

 found this to be the case within ordinary limits in 

 good instruments, there remains the question to what 

 height of mercury at what temperature a given read- 

 ing of the aneroid corresponds. For scientific pur- 

 poses it is customary to reduce the reading of the 

 mercurial barometer to the temperature of the freezing- 

 point of water, and it is often supposed that the 

 aneroid reading corresponds to that figure. But we 

 may feel pretty confident that the maker, in laying 

 down the scale, did not work in a room at freezing- 

 point. I have been accustomed to assume 15° Cent, 

 or 59° Fahr., as about the probable temperature with 

 instruments made in our climate. 



In the present case, the barometer-reading of 30'o6 

 inches at the temperature of 84° Fahr. would (neg- 

 lecting the small correction for capillarity) be reduced 

 by about fourteen-hundredths of an inch, in order to 

 give the correct figure at freezing-point ; but for com- 

 parison with an aneroid, supposed to have been laid 

 down at 59° Fahr., the correction would be a fraction 

 over seven-hundredths of an inch. As a matter of 

 fact, my aneroid marked at four p.m. 29.89 inches, or, 

 allowing for the correction, just one-tenth of an inch 

 less than the ship's mercurial barometer, and, as I 

 believe, was more nearly correct. 



As the sun was declining on the evening of July 30, 

 we sighted the remarkable island of Fernando 

 Noronha. It lies about four degrees south of the 

 equator, and more than two hundred miles from the 



