We were still at an elevation of 1,100 ft. The water was almost stagnant, and was evidently being held up by some obstacle. I feared that we should soon encounter nasty rapids. Watching the sky, I was generally able to foretell what was ahead of us in the river. In fact, a pretty mackerel sky, particularly to the north-west, showed me that the water of our river must be breaking up considerably, either in rapids or waterfalls, in order to produce sufficient moisture in the air to cause the accumulation of those cloudlets. I always noticed that wherever there were heavy rapids farther down clouds of more or less magnitude formed directly above them at a comparatively low elevation, and remained there owing to the perfect stillness of the air.

On the night of July 14th the cold was felt intensely by my men, the thermometer actually showing a minimum of 38° F.

During the night my men had a great excitement. A large pachyderm, an anta (Tapirus Americanus) inquisitively came in the midst of our camp. It was evidently as much astonished at seeing us as we were in discovering its presence. My men had been firing their cartridges away during the day at rocks, at fish in the river, and so on, so that when their rifles were really needed the magazines were all empty, and gave the anta plenty of time to hop away gracefully into the darkness of the forest.

I had given orders to them to keep watch all night, as a precaution against an attack from the Indians, but my orders were, as usual, disobeyed. Personally, I took the first watch every night, sitting up till 2 a.m., which time I occupied in writing up my notes, working out computations of astronomical observations, classifying the botanical and geological specimens collected during the day, and replenishing my cameras with new plates.

My men had eaten up all the supply of beans (feijao) I had purchased at Diamantino, and therefore even the cook could not be kept awake during the night. The first rubber collector I had picked up when coming down the Arinos was now our cook, and diabolical indeed was his cuisine. Several times already his life had been in danger from the angry attacks of his companions, the quantities of pepper he sprinkled on everything he cooked causing us all to cough sometimes for half-hours at a time. He was very fond of pepper himself, and could not understand why none of us liked it.

During the night we still had a mackerel sky, covering one-third of the sky vault, and a clear triangle of mist, the apex of which was to the west, extending towards the east, close upon the horizon line. When we left in the morning at 7.30, we had chapada and campos on the right bank and forest on the other side. We had gone some 8½ kil. from our camp when we came to a hill range, 75 ft. high, on the right bank, encircling the river with its thickly wooded slopes. There was a tributary 25 m. wide, a most beautiful stream, on the right bank. It came from 70° b.m. Its water was deliciously clear. Where it entered the Arinos it had deposited a bank of crystals and marble pebbles—yellow, red, and white—which in the dazzling sun shone with great brilliancy at the bottom of the river. Numberless rubber trees were to be seen at that spot on the banks of the Arinos, and also on those of this new important tributary.

Two kilometres farther, where the Arinos was 280 m. wide, it looked just like a big lake of stagnant water. The country was quite open on the left side, first chapada, then campos.

By 9.30 a.m. we had a most wonderful display of clouds and radiations of what looked like so many mares' tales from the W.S.W. The river at that point flowed for 1 kil. in a direction due south. We came to a basin 300 m. across with a spit of white sand on the north-west side. In this basin was an island—NattalÌ Island—200 m. long, 20 m. wide, 10 ft. above water, with a fine beach of sand and gravel on the south side. Gravel mounds were innumerable in the centre of this stream.

After we had gone some 8 kil. farther down my men shot an ariranha. They had a belief that these ariranhas would easily kill a man in the water. As we have already seen, they certainly had a great craving for blood and were always brave in attacking. My men called them "water leopards." In fact, the head of the ariranha was not unlike the head of a cat or a leopard. Although shot through the body two or three times, the ariranha actually came thrice to the attack of the canoe—so that my men were able to seize it by the tail and pull it inside the canoe while it was in a dying condition.

Sixteen kilometres farther down we came to another beautiful tributary with delightfully clear water, 6 m. wide where it met the Arinos. One hundred metres lower down another little tributary, only 4 m. wide, also on the right bank, joined our stream. The first tributary seemed to come from the north-east. At the mouth of this tributary was a spot which would have made a lovely halting place, but as it was too early in the day we reluctantly went on in a north-westerly direction, first for 4 kil., then north-east for 5 kil., passing through a large basin 300 m. wide, containing two islets, then passing charming sand-beaches, and farther on another tributary, 8 m. wide, on the left of us, also with deliciously clear water. When we proceeded on our journey after lunch we found big rocks more frequent in the stream, and went over a field of great boulders just under the surface of the water that stretched half-way across the shallow river.

Eight kilometres from our halting-place we came to an extensive stony place with a strong rapid. One kilometre beyond, a small tributary flowed into the Arinos from the left side. On the left side we had a red and brilliant yellow bank 70 ft. high, part of a small range of hills which turned the river from N.N.W. to N.N.E. Another small tributary 2 m. wide was seen on the left side. Then, 4 kil. farther on, another tributary, also 2 m. wide, and also on the left side, came from the south-west. Three thousand six hundred metres beyond this, we entered a basin 320 m. wide with an island 150 m. long, including its gravel spit. Three more islands were seen a little way beyond—Meraud, Tanis, and Loel Islands, Meraud being the largest. Another island was on the left of the river, leaving a passage 50 m. wide on its west side. The group of islands was of alluvial formation with deposits of gravel below.

The river in that region was too beautiful for words. The foliage of the thick heavy forest on both sides was densely green, the banks most tidy, and running in an almost straight line for 10,000 m. During all that distance the stream was 300 m. wide, and its speckless water reflected with marvellous definition each leaf and branch against the background of deep green. Neat gravel banks occurred frequently in the shallow water.

Some 300 m. down this long straight stretch of river a tributary 8 m. wide, coming from 210° b.m., threw itself into the Arinos. Strong eddies were formed, as many rocks were strewn in the centre of the stream.

One kilometre farther a conglomerate mass of granite and yellow and red lava, with impurities embedded in it, emerged just above the water in the centre of the stream.

Another streamlet, 2 m. wide, and of wonderfully limpid water, joined the Arinos on the right side. It came from the north-east. Then another little streamlet was seen on the left side. At the end of 10 kil., where the river made a wide angle from 330° b.m. to 350° b.m., and another straight line of 4,000 m. stretched in front of us, we beheld a huge submerged bank of sharp volcanic conglomerate rock. In fact, we unexpectedly almost ran into it. Had we done so at the rate at which we were travelling, our canoe would certainly have been smashed to pieces against the sharp-edged fractured rock—just as sharp at the angles as the blades of knives.

Where the river turned once more from 350° b.m. to 320° b.m. another small tributary appeared on the right bank, and there a lot of handsome mate trees (Ilex paraguayensis) seemed to flourish, and were certainly pretty to look at.

Farther down we again came to chapada on the left bank and heavy foliaged forest with a certain number of rubber trees on the right bank. The left bank, where it described a great sweeping circle, was low and sandy, some 12 ft. above the level of the river. Only a thin fringe of low trees grew there on the edge of the water.

Six kilometres from the last tributary on the right bank another streamlet, 3 m. wide, coming from the S.S.W., cut its way through the left bank. Two thousand five hundred metres farther on another tributary 20 m. wide—a deliciously beautiful stream—flowed gracefully into the Arinos on the right side from the north-east.

We made our camp at the junction of the two streams. The camp was extremely bad. It was already late in the evening and we could find no other suitable spot. We had gone that day 83 kils. I was quite satisfied with the progress we had made during the last few days. During the evening I made an excursion on foot along the tributary river to the north-east for several kilometres, but I found nothing of particular interest.

During the night we received another visit from an anta, but the pachyderm again escaped before my men had time to kill it. We heard cries of Indians in the distance. My men were in a great state of mind for fear we should be attacked. I sat up the entire night in order to be ready in case of emergency.

I took that opportunity of computing and checking many of the astronomical observations I had taken, and developing a great number of photographic glass plates.

In my experience I have found that the fears people have of spoiling negatives unless one is shut up in an absolutely dark room are quite exaggerated. On that particular occasion, for instance, and on many previous and subsequent occasions, I developed the glass plates—and I think with satisfactory results—out in the open, with merely the fly-leaf of the tent sheltering me overhead so as not to have the direct rays of the stars shining upon the photographic plates. Indeed, there was light enough coming in around the tent for me to see quite plainly what was going on outside. I simply covered up the developing trays as an extra precaution, and seldom—in fact, never—spoiled a negative in process of development.

I also found developing tanks quite serviceable when a great number of negatives had to be developed quickly. The red lamp necessary for photographic work was invariably a great nuisance. I do not believe that a compact, practical dark-room lamp has yet been invented which is really serviceable to an explorer. If it is a candle lamp the candle melts quickly in those hot countries, producing an extra large flame which generally cracks the red glass, and makes so much smoke that the upper aperture becomes blocked and puts the light out when you happen to be at the most crucial point of your work.

The oil lanterns would be better, were it not for the difficulty and messy nuisance of carrying and re-filling the lamp each time with oil. Electric lights, which are the only practical ones, of course are out of the question when you have to be away for a year or a year and a half, the storage batteries getting damaged easily by damp and the innumerable accidents which you have when exploring.

The greatest care had to be used in repacking the developed glass plates. I owe to the care I took of them that I was able to bring back 800 excellent negatives out of 800 glass plates exposed.

The night was a little warmer than usual on July 15th—minimum 53° F. There was a heavy mist over the river when we rose in the morning, and we had to delay our departure until 7.30 a.m. When the mist began to rise it hung about in beautiful curves converging to a common radiating centre to the west.

During the night I had noticed a weird lunar effect—a perfect cross of immense proportions intersecting the crescent moon, which had a radiating halo surrounding it.

Four thousand metres from our camp we came to a tributary 3 m. wide on the left side of the river. It came from the W.S.W. Near this a streamlet 1 m. wide entered the Arinos on the right side, and another streamlet of equal size farther down on the left bank. There was fairly thin forest on both sides as we went on, kilometre after kilometre, the water of the river being almost stagnant in that part and heavy to paddle along.

Five hundred metres down the straight stretch of river, 4,000 m. long, we came to another charming affluent, 10 m. wide, coming from the E.S.E. Farther on, another tributary 2 m. wide entered the Arinos on the left side, and formed a shallow bank of gravel extending half-way across the stream.

The Result of Half an Hour's Fishing on the Arinos-Juruena.

The Result of Half an Hour's Fishing on the Arinos-Juruena.

As I have stated elsewhere, the mentality of Brazilians was somewhat difficult to understand by people of any other nation. They did everything the wrong way, according to our notions. I had been worried a great deal, the reader may remember, at the most unpractical way in which my men loaded the animals when I had my caravan of mules and horses. I had been more than amazed at Brazilian ideas of architecture, sculpture, painting and music. I had on many occasions been dumbfounded at their ideas of honour and truthfulness. Now once more I was sickly amused—I had by then ceased to be amazed or dumbfounded or angry—at the way my men daily packed the baggage in the canoe. The baggage was naturally taken out of the canoe every night when we made our camp, for the canoe leaked so badly that when we arrived anywhere and halted we had to beach her, or else, where this was not possible, we found her in the morning almost entirely submerged. Naturally we invariably selected shallow places where we could bale the water out and float her again.

Returning to the baggage: the men every morning insisted on loading the canoe in front, where the four men were situated paddling, and the three dogs of the expedition were also accommodated. I sat in the centre of the canoe, and Alcides at the helm naturally stood in the stern. The man whose incessant daily occupation it was to bale out the water of course had to be with the group of four men in the bow, since, the canoe being so heavily weighted at that end, the water found its way down there.

Now, loading the canoe in such a fashion, at the bow, had the double drawback of causing a greater resistance against the water, and therefore nearly doubling the work of the men in paddling. Then again, when we ran aground or struck a rock, the impact was more severe on the canoe—not to speak of the difficulty of getting her off again. The steering, too, was also much more difficult with the stern of the canoe so far out of the water.

I pointed out the mistake to my men, but it was no use arguing, and they refused to follow my advice. Like all ignorant people, they thought they knew everything better than anybody else, and as, in a way, they were the chief sufferers for their own conceit, I thought I would avoid unpleasantness and let them do things their own way as long as we kept going forward on our journey.

Alcides, too, who by now had become imbued with the idea that he was as good a navigator as Christopher Columbus or Vasco da Gama, had the strangest notions of navigation. He never avoided grounding the canoe on every bank he saw; he never avoided dashing the canoe into every rock which stood or did not stand in our way. I never could understand exactly why he did that, except for the mischievous pleasure he derived from giving the men who were sitting at the other end of the canoe a violent bump, which often rolled them over altogether.

When we left Goyaz my men insisted on purchasing life-belts in case we should be travelling by water. As only one of the Goyaz men could swim, I had gladly given them the money to purchase those articles. On our first day of navigation the men amused me very much, as they all appeared garbed in their life-belts, as if we had been going to the rescue of a stranded ship in a tempest. I laughed heartily at the sight. The intense heat of the sun made the heavy cork belts so uncomfortable for them, that they discarded them when they saw that the canoe would actually float on the water, and packed them away inside a wooden box, which they then screwed down tight. The belts remained in that box most of the time, except one day when a man put one on, as I had given him instructions to go some way off in the centre of the stream where the current was rather swift. By misadventure he lost his footing, and had we not been quick in going to his rescue he certainly would have been drowned.

We tested the life-belts, and I found that not only would they not float after they had been a minute or two in the water, but they became so heavy when soaked with moisture that they would have dragged to the bottom even a fair swimmer. They were evidently old discarded ship belts. The cork, enclosed in a canvas cover, had got decomposed and pulverized, and therefore rendered useless.

As we are referring to the strange ways of looking at things by different nations, I might as well include the endless arguments I had with my men in selecting our camps. I naturally always selected the cleanest spots with a flat ground, so that the tents could be pitched satisfactorily without extra trouble, where there was little vegetation, and where the water was good. My men always quarrelled over this, and insisted on stopping in the filthiest places, either where some trees, rotted away, had fallen down, where the vegetation on the edge of the river needed cutting, and where the ground had to be levelled before I could pitch my camp bed. They always preferred sleeping under the stifling vegetation to where there was an open space and we had the clear sky over us.

They all slept in hammocks—the favourite resting arrangement of the Brazilian—to my mind the most uncomfortable and absurd fashion of resting, especially in tropical regions. First of all, it is almost an impossibility to assume a perfectly horizontal position for your entire body, except—if you are an expert—diagonally; then there is always a certain amount of swing and you are likely to tumble over at any moment; you can never keep the blankets in position, and you expose your entire body to the stings of the mosquitoes, flies and other insects, and of the ants which crawl into your hammock by hundreds from the trees in which they swarm. It was not uncommon when we camped to hear during the night a crash, followed immediately after by oaths. The tree to which one of the hammocks had been fastened had suddenly broken and let the man down with a bump. Then again, the mischievous ants took the greatest delight during the night in cutting the strings of the hammocks, and on several occasions my followers had nasty falls. Yet the Brazilians swear by hammocks.

Another stream 2 m. wide, coming from the north, entered the Arinos on the right bank. A number of ariranhas, attracted by the vivid red of the British flag which was flying at the stern of the canoe, followed us for some time and came courageously to the attack, showing their teeth fiercely at us and snarling frantically. Entire families of those delightful little creatures were seen, and they invariably gave us a similar hearty greeting. They followed us sometimes for hundreds and hundreds of metres, and became most excited when I took the flag and waved it at them, and sometimes placed it near the water in order to drive them frantic.

We now had most beautiful forest on both sides. A stream 5 m. wide joined the Arinos on the left side from the west, forming a charming little waterfall as it entered the main stream. A little farther on the right was another streamlet, coming from the south-east. Generally, as in this case, when we reached tributary streams of any importance, gravel banks extended and blocked a great part of, sometimes even half, the main stream. A picturesque stream, 8 m. wide, coming from the north-east, was then reached on the right side. It flowed through a rocky gate. Five or six kilometres farther on a tiny streamlet dribbled into the Arinos, and also another, 1 m. wide, on the left bank.

At noon that day the sky was extraordinarily interesting. From the north-west extended a wonderful succession of loop coils of transparent mist, giving the sky the appearance of a peacock's extended tail.

Just before we halted for lunch we came to a charming streamlet of delicious water, 2 m. wide, on the right bank.

The days were getting warmer as we advanced farther north. It was hot work sitting in the sun—105° F. that day—to take observations for latitude and longitude. In the shade the thermometer registered 89° F. Lat. 12° 21'·3 S.; long. 57° 16' W.

After lunch, 2½ kil. from our camp, we passed on the left bank a delightful tributary coming from the W.S.W. Its mouth was 8 m. wide, and poured forth waters of the most beautiful emerald green.

Five hundred metres farther down another large tributary, 30 m. wide, coming from the north-east, was observed on the right bank. Farther still, the river formed a large basin 300 m. wide. Lovely forest flourished round the sweeping curve of the basin. There was simply a solid mass of marvellously fresh foliage, with hardly a break through which, it seemed, a human being could pass. In that particular part the leaves came right down to the water, but there was no reason to suppose that they grew equally low inland.

The stream, which was 250 m. broad, showed farther on an immense bank of gravel 700 m. long, which rose above the surface in the shape of two long islands—one 300 m., the other 400 m. in length.

We felt the heat considerably going down the river, as we were always in the sun in the centre of the stream, with a temperature seldom less than 105° F. Especially where thick forest was on both sides of us, there seemed to be no air close to the water. When we came to patches of chapada and open country we could breathe a little better. Several were the tributary streamlets to which we came that afternoon. First we saw one rivulet, 1 m. wide, on the right bank, then 13 kil. 500 m. farther on another affluent, 3 m. wide, coming from the north-east, also on the right bank; then 1,500 m. farther a rivulet ½ m. wide, coming from the south-west (left bank); then 4,500 m. farther a charming stream, 6 m. wide, coming from the north, and meeting with the Arinos near an extensive stony place with shallow and troublesome water. Strong eddies formed at that spot. One more streamlet, 1 m. wide, was reached that day on the right. It came from the north-east.

The river had that day flowed almost continuously in directions varying from north-west to north, barring two sections where its course had been 10° east of north.

After passing the last tributary the river described a sweeping curve, gradually turning so far back as to flow in a south-westerly (240° b.m.) direction.

There was there shallow water with gravel banks in the centre of the stream. Curiously enough, we did not notice so much rubber close to the river in that region, but in an excursion a short distance from the water we came upon Siphonia elastica trees, not only along the Arinos but also along the tributaries.

We halted that day at sunset, having gone 73 kil. 400 m.; which, although much less than the previous days, was still fair going for us.