632 
APPENDIX. 
We found the barometer in this passage an 'invaluable instrument ; upon no occasion 
did it deceive us. In passing these latitudes my attention was drawn to the changes in the 
temperature of the water, which I usually found to precede a shift of wind from south to north, 
and vice versa, even before that of the temperature of the air. I subjoin a short statement 
of these changes for the satisfaction of such as may feel interested in them. 
On the 29th of August, at eight a. m. the temperature of the surface was 58°, the 
weather moderate and cloudy, and the wind "W. N. W. ; from this time to midnight it 
gradually fell until it stood at 48°. The wind now increased, and the next morning 
shifted to S. W. and S. S. W., and blew fresh gales : the breeze continuing, on the 31st 
the temperature of the surface underwent a further fall of 34°; and we had hard 
squalls, with hail and sleet. It afterwards fluctuated four days between 46° and 494°, 
during which time the winds were variable from S. S. W. to N. IV. by N., and E. N. E. 
tJie weather for the most part moderate and cloudy; but on the 5th (Sept.) the tem- 
perature (always alluding to that of the surface of the sea) rose to 53°, and the wind came 
from N. E. by N. and N., but light. The next day it shifted to S. E. by S., and the tem- 
perature rather decreased, but the breezes were light. On the afternoon of the 7th, after a 
calm, during which it remained at 50°, there was a decrease of 8° ; and thirty-six hours 
after a gale from S. by E. suddenly arose. During the five following days it was nearly 
stationary, at the temperature of 394°, and the wind was variable from W. N. W., S. W., 
and W. blowing hard. From noon on the 12th to four a. m. on the 13th it fell to 3fi°, and 
that night we had a gale at W. by S.; which continued all the next day. At night there 
was a further decrease of 4°, when the wind veered to S. by W., and blew strong gales. The 
temperature kept down at 35° until midnight of the 15th, when it rose 5°; and the l6th, at 
four A. M., the wind changed to W. N. W. and N.VV. by W. The temperature, however, 
soon decreased again 4", and at nine a. m., the following day tl>e wind came from S. W. by S. 
and S. S. E., where it continued, and the temperature remained nearly stationary until 
we made Cape Horn, when it rose to 42°. 
It would, perhaps, be too hazardous to assert upon such short experience that these 
changes are the forerunners of shifts of wind, though I found similar variations attend the 
southerly gales off Spitzbergen, where we had always indication of their approach by the in- 
crease of the temperature of the sea*. I am, however, persuaded that, like the barometer, it 
speaks a language which, though at times not the most intelligible, may nevertheless often 
prove useful. 
HOME PASSAGE FROM COQUIMBO TO RIO JANEIRO. 
June 2td lo July 21 st, 1828. 
This passage was considerably lengthened by not getting to the westward in low 
latitudes. From the time of leaving Coquimbo there was a difficulty in making progress in 
that direction, and we could scarcely weather Massa Fuera. From here the weather became 
boisterous, the breeze generally beginning atW. N.W., and ending in a moderate gale at S.W. ; 
then backing again, and in the course of the twenty-four hours finishing at S. W., blowing 
* See also p. 237 of this work. 
