EXPERIMENTS. 453 



per minute, making 1 pint (16 ounces) in every 24 hours. I then 

 cut the branch so much that, during the day, it broke ; but they 

 still went on at the rate of a drop every 5 seconds, while another 

 colony on a branch of the same tree gave a drop every 17 seconds 

 only, or at the rate of about 10 ounces 4| drams in 24 hours. I 

 finally cut off the branch ; but this was too much for their patience, 

 for they immediately decamped, as insects will do from either a 

 dead branch or a dead animal, which Indian hunters soon know, 

 when they sit down on a recently-killed bear. The presence of 

 greater moisture in the air increased the power of these distillers : 

 the period of greatest activity was in the morning, when the air 

 and every tiling else was charged with dew. 



Having but one day left for experiment, I found again that 

 another colony on a branch denuded in the same way yielded a 

 drop every 2 seconds, or 4 pints 10 ounces in 24 hours, while a 

 colony on a branch untouched yielded a drop every 11 seconds, 

 or 16 ounces 2^f drams in 24 hours. I regretted somewhat the 

 want of time to institute another experiment, namely, to cut a 

 branch and place it in water, so as to keep it in life, and then 

 observe if there was any diminution of the quantity of water in 

 the vessel. This alone was wanting to make it certain that they 

 draw water from the atmosphere. I imagine that they have some 

 power of which we are not aware, besides that nervous influence 

 which causes constant motion to our own involuntary muscles, 

 the power of life-long action without fatigue. The reader will re- 

 member, in connection with this insect, the case of the ants al- 

 ready mentioned. 



December 14th. Both myself and men having recovered from 

 severe attacks of fever, we left the hospitable residence of Mr. 

 Canto with a deep sense of his kindness to us all, and proceeded 

 on our way to Ambaca. (Lat. 9° 16" 35" S., long. 15° 23" E.) 



Frequent rains had fallen in October and November, which 

 were nearly always accompanied with thunder. Occasionally the 

 quantity of moisture in the atmosphere is greatly increased with- 

 out any visible cause: this imparts a sensation of considerable 

 cold, though the thermometer exhibits no fall of the mercury. The 

 greater humidity in the air, affording a better conducting medium 

 for the radiation of heat from the body, is as dangerous as a sud- 

 den fall of the thermometer : it causes considerable disease among 



