GREAT HEAT 



317 



in his undeviating course, and soon after penetrated a 

 dense swampy thicket, where of course I did not choose 

 to follow him. 



I suffered terribly from the heat and mosquitoes as the 

 river sank with the increasing dryness of the season, 

 although I made an awning of the sails to work under, and 

 slept at night in the open air with my hammock slung 

 between the masts. But there was no rest in any part ; 

 the canoe descended deeper and deeper into the guUey, 

 through which the river flows between high clayey banks, 

 as the water subsided, and with the glowing sun overhead 

 we felt at mid-day as if in a furnace. I could bear scarcely 

 any clothes in the daytime between eleven in the morning 

 and five in the afternoon, wearing nothing but loose and 

 thin cotton trousers and a light straw hat, and could not 

 be accommodated in Joao Aracu's house, as it was a small 

 one and full of noisy children. One night we had a terrific 

 storm. The heat in the afternoon had been greater than 

 ever, and at sunset the sky had a brassy glare : the black 

 patches of cloud which floated in it, being lighted up now 

 and then by flashes of sheet lightning. The mosquitoes 

 at night were more than usually troublesome, and I had 

 just sunk exhausted into a doze towards the early hours of 

 morning when the storm began ; a complete deluge of 

 rain with incessant lightning and rattling explosions of 

 thunder. It lasted for eight hours ; the gray dawn open- 

 ing amidst the crash of the tempest. The rain trickled 

 through the seams of the cabin roof on to my collections, 

 the late hot weather having \varped the boards, and it gave 

 me immense trouble to secure them in the midst of the 

 confusion. Altogether I had a bad night of it, but what 

 with storms, heat, mosquitoes, hunger, and, towards the 

 last, ill health, I seldom had a good night's rest on the 

 Cupari. 



A small creek traversed the forest behind Joao Aracu*s 

 house, and entered the river a few yards from our anchor- 

 ing place. I used to cross it twice a day, on going and 

 returning from my hunting ground. One day early in 

 September, I noticed that the water was two or three 

 inches higher in the afternoon than it had been in the 

 morning. This phenomenon was repeated the next day, 

 and in fact daily, until the creek became dry with the con- 

 tinued subsidence of the Cupari, the time of rising shifting 

 a little from day to day. I pointed out the circumstance 



