116 MANGROVES. 



ited by mestizo fishermen, the livid tint of whom, 

 together with the miserable appearance of their 

 children, gave indication of the unhealthy nature of 

 the coast. The sea is so shallow that one cannot go 

 ashore in the smallest boat without wading. The 

 woods come nearly to the beach, which is covered 

 with mangroves, avicennias, manchineel-trees, and 

 Suriana maritima, called by the natives romero de la 

 mar. Here, as elsewhere, the insalubrity of the air 

 is attributed to the exhalations from the first of these 

 plants. A faint and sickly smell was perceived, re- 

 sembling that of the galleries of deserted mines. 

 The temperature rose to 93-2°, and the water along 

 the whole coast acquired a yellowish-brown tint 

 wherever it was in contact with these trees. 



Struck by this phenomenon, Humboldt gathered 

 a considerable quantity of branches and roots, with 

 the view of making experiments on the mangrove 

 upon his arrival at Caraccas. The infusion in warm 

 water was of a brown colour, and had an astringent 

 taste. It contained extractive matter and tannin. 

 When kept in contact with atmospheric air under a 

 glass jar for twelve days, the purity of the latter 

 was not perceptibly affected. The Avood and roots 

 placed under water were exposed to the rays of the 

 sun. Bubbles of air were disengaged, which at the 

 end of ten days amounted to a volume of 40 cubic 

 inches. These consisted of azote and carbonic acid, 

 with a trace of oxygen. Lastly, the same substances 

 thoroughly wetted were enclosed with a given vol- 

 ume of atmospheric air in a phial. The whole of 

 the oxygen disappeared. These experiments led 

 him to think that it is the moistened bark and fibre 

 that act upon the atmosphere, and not the brownish 

 water which formed a distinct belt along the coast. 

 Many travellers attribute the smell perceived among 

 mangroves to the disengagement of sulphuretted 

 hydrogen, but no appearance of this kind was ob- 

 served in the coiu-se of tliese investigations. 



