212 REACTION OF THE INTERIOR OF THE EARTH 



opening of from 21^- to 30 inches diameter, surrounded 

 by a little mud wall. The gas rises with great vehe- 

 mence, as at Taman, in bubbles, each of which, accord- 

 ing to my measurements made in graduated glass vessels, 

 contains from 10 to 12 (French) cubic inches. The 

 upper part of the funnel is filled with water resting on 

 a thick mud covering. Adjacent cones do not have 

 simultaneous eruptions, but in each single cone there 

 may be remarked a certain degree of regularity in the 

 times of eruption. Bonpland and I counted pretty re- 

 gularly five eruptions every two minutes. When one 

 stoops over the little crater opening, one can perceive, 

 generally about twenty seconds before each eruption, a 

 hollow noise in the interior of the earth deep below the 

 base of the cone. In the gas which rose, and which was 

 on two occasions collected with great care, a lighted 

 very thin wax taper was instantly extinguished, as was 

 also a burning splinter of Bombax Ceiba. The gas 

 could not be made to ignite. It did not render lime- 

 water turbid ; no absorption took place. When tested 

 by nitric oxide gas for oxygen, it showed no trace of the 

 latter ; in another experiment, in which the gas of the 

 volcancitos was kept for many hours in a small bell-glass 

 turned down in water, it showed rather more than one 

 per cent, of oxygen, which was probably derived from 

 the water, and had become accidentally mixed with it. 



According to these results of analysis, I then stated 

 that I considered (not altogether incorrectly) the gas 

 of the volcancitos of Turbaco to be nitrogen, which 

 might be mixed with a small quantity of hydrogen. * I 

 expressed at the same time in my journal regret, that 

 in the then state of chemical knowledge (in April 1801) 



