NOTES ON BRAZIL. 



385 



larger trunk was seventy-four feet long, near fhe root five feet 

 round, and at the other end four ; both Avere without bark, seemed to 

 have lost, by comparing them with their respective roots, about half 

 their circumference, and would have yielded timber to a greater length ; 

 they were perfectly straight, and of that kind of forest-wood which 

 throws out no buttresses at the root. It is astonishing that such trees 

 stand so firmly as they do, for the head is generally very close, loaded 

 with plants, and exposed above the forest to the full power of tlie 

 wind, acting upon a long lever. 



While measuring these trunks, I was stung by an insect, whose 

 nest was a cone, with the apex attached to a thorn. It contained 

 thirteen cells, whose mouths were downwards, the larvee being suspended 

 within them : these were vigorous, but not sufficiently advanced, I 

 presume, to show, to a more skilful inquirer than myself, what class 

 they belonged to. I caught also a moth, which I should be glad to 

 describe scientifically, for it seemed remarkable both for its size and 

 beauty. It was an inch and a quarter long, its colour a silvery white, 

 with regular and finely formed patches of a light delicate drab or dove 

 colour ; the form of the insect was extremely pleasing, and its attire pecu- 

 liarl}'' rich. — This morning, before we set off, the thermometer was 54°, 

 at noon 84", and at seven in the evening 75". 



Early the next morning we proceeded along the same valley, and soon 

 found it close considerably ; at about two miles distance the volume of 

 water in the river was greatly increased, probably by the reception of a 

 current from the Eastward, while it was out of our sight, making a 

 short detour round a hill to our right ; yet .the stream still flowed in 

 a rather even bed, A little farther on the Dell became deeper and 

 narrower, and the river tiunbled down steep declivities among masses of 

 stone ; the road was merely a narrow mule-track, cut out of the side of 

 a mountain, and was frequently three, and sometimes five hundred feet 

 above the level of the stream ; once it descended to the water's edge. 

 The few rills which crossed the road were so small and short as to con- 

 vince us that we were travelling along the western edge of the basin 



3 c 



