118 HORTICULTURAL DIFflCUTIES. 



paintings ; and that small cakes of this substance were 

 sold at the great market of Tenochtitlan. But a colouring 

 matter, chemically identical, may be extracted from plants 

 belonging to neighbouring genera; and I should not at 

 present venture to affirm that the native indigoferse of 

 America do not furnish some generic difference from the 

 Indigofera anil, and the Indigofera argentea of the Old 

 World. In the coffee-trees of both hemispheres this 

 difference has been observed. 



Here, as at the Rio Negro, the humidity of the air, and 

 the consequent abundance of insects, are obstacles almost 

 invincible to new cultivation. Everywhere you meet with 

 those large ants that march in close bands, and direct their 

 attacks the more readily on cultivated plants, because they 

 are herbaceous and succulent, whilst the forests of these 

 countries afford only plants with woody stalks. If a mis- 

 sionary wishes to cultivate salad, or any culinary plant of 

 Europe, he is compelled as it were to suspend his garden in 

 the air. He nils an old boat with good mould, and, having 

 sown the seed, suspends it four feet above the ground with 

 cords of the chiquichiqui palm-tree; but most frequently 

 places it on a slight scaffolding. This protects the young 

 plants from weeds, worms, and those ants which pursue their 

 migration in a right line, and, not knowing what vegetates 

 above them, seldom turn from their course to climb up 

 stakes that are stripped of their bark. I mention this cir- 

 cumstance to prove how difficult, within the tropics, on the 

 banks of great rivers, are the first attempts of man to fctvoro- 

 priate to himself a little spot of earth in that vast domain 

 of nature, invaded by animals y and covered by spontaneous 

 plants. 



During the night of the 13th of May, I obtained some 

 observations of the stars, unfortunately the last at the Cas- 

 siquiare. The latitude of Mandavaca is 2 4' 7" ; its longi- 

 tude, according to the chronometer, 69 27'. I found the 

 magnetic dip 25'25 C (cent, div.), showing that it had increased 

 considerably from the fort of San Carlos. Yet the sur- 

 rounding rocks are of the same granite, mixed with a little 

 hornblende, which we had found at Javita, and which as- 

 sumes a syenitic aspect. "We left Mandavaca at half-past 

 two in the morning. After six hours' voyage, we passed on 



