248 TRAVELS ON THE RIO NEGRO. [March,
though it had been seen in its cage on the preceding day. I thus lost perhaps my only chance of obtaining a much-desired and probably undescribed river turtle, as the time of egg-laying was past, and they had now retired into the lakes, and become very scarce and difficult to be met with.
As my Indians were here doing nothing, I sent three of them with Sebastiao up the Codiari, with beads, hooks, mirrors, etc., to buy monkeys, parrots, or whatever else they could meet with, as well as some farinha, which I did not wish to be in want of again. I sent them with instructions to go for five or six days, in order to reach the last stitio, and purchase all that was to be had. In two days, however, they returned, having been no further than Philippe had gone, Sebastiao saying that his companions would not go on. He brought me some parrots and small birds, bows, bird-skins, and more farinha than my canoe would carry, all purchased very dearly, judging by the remnant of articles brought back.
Being now in a part of the country that no European traveller had ever before visited, I exceedingly regretted my want of instruments to determine the latitude, longitude, and height above the sea. The two last I had no means whatever of ascertaining, having broken my boiling-point thermometer, and lost my smaller one, without having been able to replace either. I once thought of sealing up a flask of air, by accurately weighing which on my return, the density of air at that particular time would be obtained, and the height at which a barometer would have stood might be deduced. But, besides that this would only give a result equal to that of a single barometer observation, there were insuperable difficulties in the way of sealing up the bottle, for whether sealing-wax or pitch were used, or even should the bottle be hermetically sealed, heat must be applied, and at the moment of application would, of course, rarefy the air within the bottle, and so produce in such a delicate operation very erroneous results. My observations, however, on the heights of the falls we passed, would give their sum as about two hundred and fifty feet ; now if we add fifty for the fall of the river between them, we shall obtain three hundred feet, as the probable height of the point I reached above the mouth of the river ; and, as I have every reason to believe that that is not five hundred feet above the sea, we shall obtain eight hundred feet as the probable limit of
