334 NOTES OF A NATURALIST. 



from the scene of action, but, though watching atten- 

 tively, I saw no appearance of heaving of the surface 

 or any break in its regular form. The waste seemed 

 to proceed altogether from the upper surface, and the 

 emergence of the prominent objects in regular suc- 

 cession gave direct evidence to that effect. 



During the first five days of my visit the weather 

 at Petropolis was perfectly enjoyable. The tempera- 

 ture varied from about 60° Fahr. at sunrise to about 

 70° in the afternoon ; but the effect of radiation must 

 have been intense, as in an exposed situation a 

 minimum thermometer descended on one night to 

 46°, and on the next to 44°, and the dew was heavier 

 than I have ever seen it elsewhere, so that in some 

 places the quantity fallen from the leaves of the trees 

 made the ground perfectly wet in the morning. The 

 barometer varied very little, even after the weather 

 changed, and stood as nearly as possible three inches 

 lower than at Rio, showing a difference of level of 

 about 2900 feet. On the i6th of July the sky became 

 overcast, and some rain fell in the afternoon, the ther- 

 mometer rising at two p.m. to 73° Fahr., and moderate 

 rain fell on each succeeding day until the evening of 

 the 19th, but scarcely any movement of the air was 

 perceptible. There is a remarkable difference in the 

 distribution of rainfall between the part of Brazil lying 

 within about fifteen degrees of the equator and the 

 region south of that limit. At Pernambuco (south 

 lat. 8° 4'), out of an annual rainfall of about a hundred 

 and ten inches, nearly ninety inches fall during the 

 six months from March to August, and at Bahia, with 

 less total rainfall, the proportion is nearly the same. 



