Aus;. 1836. pernambuco. 593 



'^& 



at the thought of refusing such a request, or of behaving to 

 a stranger with rudeness. The channel by which we went 

 to, and returned from Ohnda, was bordered on each side by 

 mangroves, which sprang Uke a miniature forest out of the 

 greasy mud-banks. The bright green colour of these bushes 

 always reminded me of the rank grass in a churchyard : 

 both are nourished by putrid exhalations; the one speaks 

 of death past, and the other too often of death to come. 



The most curious thing which I saw in this neighbourhood, 

 was the reef that forms the harbour. It runs for a length 

 of several miles in a perfectly straight line, parallel to, and 

 not far distant from, the shore. It varies in \^idth from 

 thirty to sixty yards ; it is dry at low water, has a level 

 smooth surface, and is composed of obscurely-stratified 

 hard sandstone. Hence, at first sight, it is difficult to credit 

 that it is the work of nature, and not of art. Its utility is 

 great ; close within the inner wall there is a good depth of 

 water, and ships lie moored to old guns, which are fixed 

 in holes on its summit. A lighthouse stands on one 

 extremity, and around it the sea breaks heavily. In enter- 

 ing the harbour, a ship passes within thirty yards of this 

 point, and amidst the foam of the breakers ; close by, 

 on the other hand, are other breakers, which tlius form a 

 narrow gateway. It is ahnost fearful to behold a ship run- 

 ning, as it appears, headlong into such dangers. 



With respect to the origin of the reef,' I beheve a bar 

 composed of sand and pebbles formerly existed beneath the 

 water (a circumstance no ways improbable) when the low 

 land, on which the town now stands, was occupied by a large 

 bay; and that this bar was first consoHdated, and then 

 elevated. These two processes are of such frequent occur- 

 rence in South America, that there can be no objection to 

 using them in accounting for any remarkable structure in the 

 land. There is another and slightly different explanation 

 which possesses equal probability, namely, that a long spit 

 of sand, like some that now run parallel to parts of the 

 neighbouring coast, had its central part consolidated, and 



VOL. III. 2 Q 



