We had gone 96 kil. in four days' marching since leaving the Rio Manso. We were only a few kilometres from the Serra Azul, or Blue Mountains—truly mountains of the most vivid and purest cobalt blue I had ever seen—quite a wonderful spectacle.

We made our camp in a prairie with good grazing for our animals. Although we were at a comparatively low elevation—1,150 ft. above the sea level—the minimum temperature of the atmosphere was 56° Fahr. during the night.

On leaving camp—still proceeding north—we descended to 1,100 ft. into a lovely stretch of magnificent grass with a lagoon. The level of the water was low, as we were then at the end of the dry season. On the flat grassy land were curious semi-spherical mounds, 4 to 6 metres in diameter and from 2 to 6 ft. high. On each of these mounds were a few stunted trees. No trees whatever existed except upon these small mounds, the explanation being, I think, that the mounds had formed around the trees while these were growing, and not that the trees had grown upon the mounds. As we were getting nearer, the Serra Azul to the north was most impressive. I think that it was partly due to the bluish foliage of the vegetation upon it that the range, even close by, appeared of so vivid a blue, and also to the deep blue shadows cast by the spurs which projected, some to the south-east, others due south—that is, it will be understood, on the southern face of the range.

Thick deposits of cinders lay in the valley. On approaching an intermediate and lower range we cut our way through scrub—chiefly of sciadera trees, seldom growing to a greater height than 7 ft. The domed hills showed through the grass great blocks of volcanic rock, while at the foot of the hills could be noticed huge boulders of consolidated ashes with veins of crystals and marble. There, too, the stratification was vertical. There was lamination in some of the rock, but not in the granite blocks nor in the blocks of marble, which appeared to have been subjected to enormous heat. Some of the rock had been in a state of absolute ebullition.

Marvellous Scenery of the Central Brazilian Plateau.

Marvellous Scenery of the Central Brazilian Plateau.

"Church rock" standing in the centre.

At the spot where we crossed the range—starting our ascent from an elevation of 1,100 ft.—were immense holes, vents and cracks in the earth's crust. As we rose slightly higher among many chains of low hills, we were upon a horizontal stratum of laminated granite. Higher still we passed a semicircular hill composed of immense blocks of granite. In the centre of the semicircle was a great round hole, 30 ft. in diameter—an extinct crater. Farther on, ascending upon an inclined plane, we came to another similar semicircle—not of rock that time, but of red earth and cinders. When we reached the highest point (elev. 1,270 ft.) of the divide we had to our left huge pinnacles and pillars of rock of the most fantastic shapes, monoliths from 10 to 15 ft. high, and rocks hollowed by the action of fire. Big boulders, which had become perfectly rounded by having been shot through the air and revolved at a great speed while in a half-solid condition, were to be seen scattered all over the inclined planes of the saddle of the divide. Giant cacti grew in abundance in the interstices between rocks. Although most of the rocks were blackened outside, by chipping off the outer surface one found that they contained inside beautiful white marble or else greyish granite. The latter was striated with thin layers—not more than a quarter or half an inch thick—of crystallized matter, forming veins in the blocks or dividing two strata.

Everywhere could be noticed remarkable perforations of all sizes in the rocks, great spherical or ovoid hollows, or cylindrical tubular channels. In the ground were many volcanic vents with lips baked by fire.

On our right, a kilometre or so farther on, after having gone through an extensive stretch of red sand and lapilli, we came across three hills, the central one of which had the appearance of a cylindrical tower of masonry with windows and doors. It was a wonderful freak of nature. Under this huge tower were several caves and grottoes.

Descending upon the opposite side of the range, at an elevation of 1,200 ft. we found the dry bed of a streamlet, which flowed in a northerly direction when it did flow at all. On emerging from the wide hill mass—about 18 kil. across—we found ourselves among a lot of burity palms on the western spur of the Serra Azul. When we were actually upon them, the Blue Mountains lost their blue appearance and were more of a greyish green, owing to the vegetation which covered most of their slopes. The range was formed of three distinct terraces, the lower one being of greater height than the two upper ones. A number of low hill ranges starting from the main range branched off like spurs towards the south. The uppermost terrace of the main range was supported on a high vertical wall of red rock.

On meeting the Rio Coralzinho we skirted it for some distance through the forest, then marched among a great many domes, small and large; after which we crossed a wonderful field of huge monoliths, superposed boulders, and rocks of all kinds of fantastic shapes.

We had marched 30 kil. that day. We encamped on the River Piraputangas—a tributary on the left side of the CuyabÁ Grande River—the CuyabÁ Grande being in its turn a tributary on the right of the CuyabÁ River.

The CuyabÁ River described almost an arc of a circle—in fact, quite a semicircle—its birth taking place in the Serra Azul. Where we crossed it we were only a short distance to the west from its point of origin.

Where we had made our camp we were in a large grassy plain about six kilometres long and nearly two kilometres wide. The rainy season was fast approaching. We came in for a regular downpour during the night, accompanied by high wind, which knocked down all our tents, as the pegs would not hold in the soft, moist ground. We had a busy time endeavouring to protect the baggage. We all were absolutely soaked. The minimum temperature was 52° Fahr. In the morning, after the wind had abated and the rain had stopped, we were enveloped in thick fog.

We had descended to so low an altitude as 750 ft. above the sea level on the north side of the Serra Azul—the lowest elevation we had been at for some considerable time. We had descended altogether from the highest part of the great Central Brazilian plateau. From that point all the waters would be flowing to the north-east or north. We were, in fact, within a stone's throw—to be more accurate, within the radius of a few kilometres—of the birthplace of the Rio Novo, the head-waters of the River Arinos, of the Rio Verde (Green River), and of the several sources of the Rio S. Manoel or das Tres Barras, or Paranatinga; and not distant from the sources of the great Xingu River.

The Serra Azul, extending from west to east, was interesting geographically, not only because it marked the northern terminus of the highest terrace of the great central plateau, but also because from it or near it rose two of the greatest rivers of Central Brazil—the Xingu and the Arinos (Tapajoz), the latter the most central and important river of Brazil, crossing the entire Republic from south to north, as far as the Amazon.

On June 21st we crossed the Piraputangas (elev. 750 ft. above the sea level), where, owing to the steep banks, we had much difficulty in taking mules and baggage to the opposite side. We then proceeded across another large plain, skirting the spurs of the Serra Azul. Nine kilometres from camp we came to a stream 80 metres wide, which flowed from north-east to south-west. It had an average depth of 1½ ft. It was, I think, the CuyabÁ Grande.

It was not easy to identify those rivers, as the existing maps of that country were absolutely worthless, most of them being filled in with fancy mountains and rivers, which either did not exist at all or were sometimes hundreds of kilometres out of their position. There were frequently mistakes of two, three, and more degrees in the latitudes and longitudes even of important places. As for the tributary rivers, of which merely the mouths were known and named, they had supplied good material for the imagination of more or less artistic cartographers in order to fill in the rest of their course. Even the German map and the American maps of the International Bureau of American Republics, which were the two best, were extremely inaccurate in their representation of that region. For instance, the latter map—and nearly all the other maps—placed the Serra Azul some 180 or 200 kil. south of its actual position. The German map was some 70 kil. out. The Serra Azul could be seen from a great distance, and had been marked approximately and not by actual observations on the spot. Nor, of course, had the tributaries of the CuyabÁ been explored or even seen except at their mouths; hence their imaginary courses.

A Street of Diamantino.

A Street of Diamantino.

The Dogs of the Expedition.

The Dogs of the Expedition.

Considering how the maps of those regions had been got together, it was really wonderful that, with all their blunders, they gave as much information as they did. Unhappy, nevertheless, would be the poor traveller who relied on those maps in making a journey across the country. For instance, if you expected to come upon a certain river in one day and did not get there until after ten or fifteen days' hard marching; if you expected to find a mountain range—nearly as high as the Himalayas or at least as high as the Andes, according to the deep shading on the maps—and found instead an interminable flat plain; and if you saw on your map rivers marked navigable, and found rapids instead, in comparison with which the terrible ones of Niagara are mere child's play, you would certainly become rather sceptical of prettily-drawn maps.

On most of the maps of Brazil one saw marked to the east of the Araguaya, in the Goyaz Province, an immense range with no less a name than Cordilheira Geral la Serra do Estrondo—or "General Range of the Mountains of Noise." They were marked as the most prominent range in Brazil—quite as high as the Andes of Peru, Bolivia, and Chili; whereas, as a matter of fact, I was told on good authority that they were mere low hills, where there were any hills at all.

To come to great geographical mistakes which came under my direct observation, I found a very palpable one in the head-waters of the CuyabÁ River, which had their source to the north of the Serra Azul and not to the south, as marked on many maps, including the Brazilian official maps.

We had to our left the Serra das Pedra—"Range of Rocks"—an extraordinarily rocky range, which was crossed almost at right angles by the ChapadÃo das Porcas. We marched through a wonderful growth of palmeiras, some of the palms being as much as 30 ft. high. Buritys were innumerable along a small stream—the Rio Estivado—flowing south-west into the CubayÁ River. There were great quantities of mangabeira trees. We proceeded northward along a chapada—a capital Brazilian name which denotes a locality that is neither a forest nor a prairie. The chapada had scanty trees and scrub, but not enough to make it into a forest.

We were marching over low hills with surface deposits of sand and cinders. We gradually reached an elevation of 1,050 ft. some 18 kil. from camp, and shortly after—and only 50 ft. lower—entered a refreshing grove of giant palmeiras and buritys along the Rio das Porcas, flowing westward. There, north of the stream, we went across more clean campos, 1,700 metres wide, bounded to the north by the thickly-wooded hill-range Keboh, extending before us from east to west.

We crossed this range in the centre, during a strong gale from the south-west. The wind cleared the sky, that had been overcast and had made the atmosphere heavy. Again that afternoon, when the wind ceased, I noticed the peculiar striations in the sky—not in straight lines that time, but in great and most regular curves converging to the west.

The valley got narrower as we went along. Two twin conical hills ended the northern extremity of the range (south-east to north-west) which we had on our left—a great mass of granite blocks in the centre of the plain rising higher and higher into regular domes. The plain itself, on an incline, showed two swellings of great magnitude, the one to our right about 120 ft. higher than the plain, the elevation of which was 1,000 ft. On the west side of those two swellings was a confused mass of huge blocks of granite—of all sizes and shapes—which to all appearances had been shot up from underneath by some internal force. They were outwardly much blackened by the action of fire, but internally were of a grey tint. A little farther we were encircled by basaltic columns of great height, many of them fractured, forming a fantastic sky-line. Some resembled the spires of a cathedral; groups of others had the appearance of the ruins of an ancient fortress; others stood up like giant obelisks; while accumulations of others formed more or less regular pyramids.

After leaving that strange basin, we were once more travelling across patches of clean chapada and dirty chapada—according to the soil and quantity of moisture; then over arid campos spreading for 15 kil. without one single drop of water.

At sundown, after having gone over several undulations varying from 850 to 900 ft. above the sea level, we went over a hill slightly higher—950 ft.—with a summit of ashes, red earth, and yellow lava pellets, as well as great sheets of foliated lava.

Under a most wonderful effect of light to the west—three superposed horizontal bands of luminous yellow, violet and brilliant vermilion, over the deep cobalt mountain range in the distance—we arrived, my men being thirsty and tired, at a little rivulet. We had marched 42 kil. that day.

My men felt the cold intensely during the night—the minimum temperature was 48° Fahr., with a high, cutting wind. Yet we were at a low elevation, merely 750 ft. above the sea level. There were, as usual, moans and groans all night, more toothache and rheumatic pains and bones aching in the morning. The discontent among my men had reached a trying point. They worried me continuously to such an extent—indeed, as never in my life I had been worried before—that I was within an ace of breaking my vow of never losing my patience and calm. In my long experience of exploring I have always had to deal with the most troublesome types of men imaginable, but never with any quite so unpleasant as those I had in Brazil.

When, the next morning, I ordered them to pack the animals in order to proceed on our journey, there was an unpleasant scene approaching mutiny. They knocked things about and refused to go on. Then they sat, rifles in hand, a little way off, grumbling and grunting, with vicious expressions upon their faces. They were going to do wonderful things—they were indeed! I overheard them. One man came forward—the spokesman. The men claimed their money up to date since the last payment made to them—only a fortnight before. They all wished to go.

"Certainly," was my immediate reply. Without a moment's hesitation they were each handed over their full pay, and without giving the slightest attention to them, Alcides, who had remained faithful, and I—poor Filippe had been dragged against himself into the plot—collected all the animals and packed them. Without one look or word—as if they had not existed—I started off the troop of animals and got on my saddle to depart last. With the corner of my eye I kept a watch on them—as with men of that kind the chief danger was when you had your back turned. I had gone only a few yards when I heard some one sobbing behind my mule. As I turned round, the two outstretched hands of Filippe were handing me back the sum of money I had paid a few moments before. He was begging me to keep it safely for him. Then two more hands urged me to take back for safe keeping the wages they had just received. The faces of the owners of those hands were too comic for words: the cheeks shining with abundant tears that streamed down, the eyes red and swollen, the mouths stretched in nervous strain from ear to ear. Behind came two more men, looking as mournful as if they were being led to execution.

They all begged to be re-employed. I let them follow—on foot—for several kilometres without saying a word—struggling through the heavy marching painfully and wading across chest-deep in the streams. We crossed the RiberÃo Chabo or Guebo, 25 metres wide and 3 ft. deep, at an elevation of 730 ft., then shortly after we waded through another stream flowing south, with a zone of wonderful palmeiras along its banks. We then emerged into a magnificent plain with a barrier of low hills to the north-west. Six kilometres farther we waded across the PlanchÃo stream, 5 metres wide and 6 in. deep. Marching on horseback was delightful, the maximum temperature being only 74° Fahr. in the shade. Another stream, flowing from north to south, the PlanchÃonzinho, whose foul water was quite disgusting to drink, although beautifully limpid, was then negotiated.

I was delighted at meeting with so many streams, for there was nothing my men hated more than to get into the water. They felt very sorry for themselves, to be struggling along as best they could, following the animals like humble sheep instead of being comfortably mounted on quadrupeds. We travelled a considerable distance through campos, but owing to some baggage which had been lost we eventually had to retrace our steps as far as the PlanchÃonzinho River, on the banks of which we encamped. This was unfortunate, as the water had a sickening flavour and made even our coffee and tea taste like poison.

Misfortunes never come alone. In overhauling my baggage I discovered, to my dismay, that my men—in order to force me to go back the way we had come—had gradually thrown away most of the provisions, which should have lasted us some six to seven months longer. We had only sufficient food to last us a few days. The men confessed their misdeed. The country provided absolutely nothing to eat, and I had to face the problem of either dying of starvation or falling back on some place where we could purchase fresh provisions. It was out of the question—unless one wished to commit suicide and a quintuple murder—to endeavour to push on towards my goal, Manaos on the Amazon, some 1,600 kil. distant as the crow flies, or at least 4,000 to 5,000 kil. travelling, with possible deviations, without some of which it was not possible to travel. We could certainly not fall back on our point of departure, the terminus of the railway at Araguary, 1,596 kil. distant; nor on Goyaz, the last city we had seen, 1,116 kil. away—so that the only way to escape death was to fall back on the ancient settlement of Diamantino, the farthest village in Central Brazil, a place once established by the first Portuguese settlers of Brazil while in search of diamonds.

Diamantino was practically in the very centre of the thicker part of South America, without counting Patagonia. It was almost equidistant—roughly speaking, some 2,560 kil. as the crow flies—from Pernambuco on the Atlantic Coast to the east, Callao (Lima) in Peru on the Pacific Coast to the west, Georgetown in British Guyana to the north, and Buenos Ayres in the Argentine Republic. Although so far in the interior and almost inaccessible from the north, east, and west, Diamantino could be reached comparatively easily from the south, travelling by river up the Parana, Paraguay, and the CuyabÁ Rivers, as far as Rosario—thence by trail to Diamantino. I had heard that the place was once flourishing, but had since become almost totally abandoned. I thought that perhaps I might be able to purchase sufficient provisions to get along; and—hope being one of my everlasting good qualities—I also dreamt that perhaps I might there get fresh men.

It was indeed with a bleeding heart—when I had reached a point some 200 kil. north of the Serra Azul—that I had to alter my course, which had been practically due north, into a south-westerly direction, and endeavour to find Diamantino. My men were delighted at the prospect of seeing human beings again. We had met no one for some weeks. We made terrific marches daily in order to reach that village before the food gave out altogether.

The nights were cold—47° Fahr. being the minimum at our camp on June 23rd.

We crossed a small range of hills over a pass 930 ft. above the sea level, and found ourselves in a spacious cuvette with the usual central line of buritys and thick vegetation (elev. 900 ft.). Soaring over our heads were a number of gaviÃo caboclo (Hetorospidias meridionalis), a kind of falcon, rending the air with their unmusical shrieks.

Matto-Grosso Girl, a Mixture of Portuguese, Indian and Negro Blood.

Matto-Grosso Girl, a Mixture of Portuguese, Indian and Negro Blood.

Brazilian Child, a Mixture of Portuguese and Negro.

Brazilian Child, a Mixture of Portuguese and Negro.

After leaving the cuvette we began to ascend the Estivado Range, very steep and rocky. Near the summit we struggled through a field of great igneous boulders, chiefly upright pillars of granite and white marble. Upon the pass (elev. 1,400 ft.) was a circular depression some 300 metres in diameter, perfectly flat-bottomed and grassy. It was surrounded by cones from 80 to 100 ft. high. On the south-east side of the range—very steep—was abundant rock, whereas to the north-west side was a padding of brown earth on a gentle incline divided into terraces. Here and there pointed noses of volcanic blocks, similar to those we had found on the opposite side of the range, showed through. We went across a depression where water dripping down the mountain-side had remained stagnant, rendering that spot almost impassable. The animals sank chest-deep into slush, crashing through the thick and much-entangled growth of live and fallen bamboos.

More campos, fairly wide, were found beyond this, and great stretches of foliated slate and sandstone in strata turned over into a vertical position, and quantities of dÉbris. Then again we cut our way through a cool growth of bamboos, handsome palmeiras and akuri palms; after which we emerged into campos once more, rising gradually to an elevation of 1,550 ft. upon an undulating terrace of the second section of the Estivado range. Pulling and pushing the mules and horses over a lot of boulders and up a steep incline, we reached the highest point of the range on our route—1,800 ft. above the sea level. Again the stratification of red and grey rock in layers from 6 ins. to 1 ft. thick, standing vertically, showed what a geological commotion there must have been in those regions. The summit of the range, extending from north to south, appeared like the teeth of a saw, so broken up was it into repeated undulations. On the west side of the range we found a gentle slope of clear campos with merely a few stunted trees upon them.

Before us to the west stood high the level sky-line of a table-land, showing perfectly straight parallel strata of rock extending all along its face, but slightly undulated near the summit of the range. Otherwise its grassy slopes were quite undisturbed in their virgin smoothness.

In the distance to the north of our course was a great lagoon—the LagÕa dos Veados, "Lagoon of the Deer"—a most important point in South America, for it was there that the great Arinos (Tapajoz) River rose. The lagoon—3 kil. long and less than 1 kil. wide—had no visible outlet, but some hundreds of metres away a spring came out of the earth, forming the Rio Preto (Black River). The Rio Preto, soon joined by the Rio Novo which we had seen descending from the Serra Azul, formed the Arinos River and could certainly be considered the head-waters of that immense tributary of the Amazon.

A short distance south of Diamantino were the Sete Lagoas, or Seven Lakes—as a matter of fact, they numbered more than seven—circular pools only a few yards in diameter but extraordinarily deep, evidently of volcanic origin, and filled with water at a later time. Around their edges a remarkably luxuriant growth of buritys could be admired. A great valley extending south with a central ridge could be distinguished. On it was the meeting-place of the Rio Diamantino and the Rio do Ouro (River of Gold), which, with the Sete Lagoas, formed another most important point of South America, for it was there that the Great Paraguay or Parana River rose.

It was thus interesting to note that within almost a stone's throw rose two of the most powerful rivers of South America—one flowing due north into the Amazon, the other almost due south as far as Buenos Ayres and Montevideo, where it entered the Atlantic Ocean.

A great confusion is made on most maps between those lagoons and the actual birth-places of those important streams. The ancient Jesuits and friars had a fair idea of geography. I have in my possession a remarkable work in Italian published in Rome in 1698 by Father John Joseph of S. Teresa—a barefooted Carmelite. It is entitled The History of the Wars in the Kingdom of Brazil between the Crown of Portugal and the Republic of Holland. The book contains a number of extraordinary maps of Brazil. Those of the principal harbours give a splendid idea of the places represented. The coastline of the continent is indicated with fair accuracy. It is curious to note that the author of that book and the cartographer place the sources of the Amazon and of the River Plate in the same spot, as descending on opposite sides of a range extending from east to west—a range which does not exist, unless it was intended to represent the Central Brazilian plateau. "The River S. Francisco," Father John Joseph goes on to state, "has also its birth in the spot where the Amazon is born, but this is not sure." The cartographer, in fact, places the head-waters of that river close to the head-waters of the Amazon, and makes them flow through a large lagoon in the heart of Brazil—evidently the Great "LagÕa dos Veados" or else the "Sete Lagoas" to which reference has previously been made in this chapter. "The Rio Grande (Rio Parana, Paraguay), one of the most celebrated in Brazil," proceeds the Carmelite Father, "is born already swollen by plentiful waters (sic) in the interior of terra firma! Near its sources it forms a lagoon 20 leagues in circumference." All this is, of course, geographically wrong. The Rio S. Francisco has its birth far to the south-east in Minas Geraes, some hundreds of kilometres distant from that lagoon and several thousand from the real source of the Amazon.

Also the friar must have mistaken—evidently from information received—the sources of the Arinos for the sources of the Amazon, which are really located some 15° of longitude west. It is nevertheless curious that so far back as 1698 the existence of the lagoon should be known at all—perhaps they had heard of it from the adventurous Paulista Bandeirantes—and that they should have placed it nearly in its proper latitude and longitude on their maps. Apparently Father John Joseph was not aware of the existence of the Great Araguaya and Xingu Rivers. Having compiled his map from information, he confused those rivers into the S. Francisco River. Upon descending from the Serra into the valley we soon came to a large forest with a luxuriant edge of peroba (a word originating, I believe, from the words ipe and roba in the Tupi language), which was known in four different varieties: viz. the peroba amarella (yellow), parda (brown), revessa (knotty), and rosa (rose-coloured), technically named: Aspidosperma polyneuron M. Arg., Aspidosperma leucomelum Warmg, Aspidosperma sp., Aspidosperma dasycarpon A.

Then there were also plentiful garabu and other tall trees. Before getting to the edge of the forest I noticed among the rocks some beautiful specimens of the apita cactus, 10 ft. and more in height, in appearance not unlike giant artichokes.

Near its beginning, where it was 3 metres wide and 6 in. deep, we crossed the Estivado River, which with a group of other streamlets may share the honour of being one of the sources of the Arinos. It flowed in a north-westerly direction.

We were pushing on for all we were worth, for we had come to the end of our food. Up and down we went over a troublesome series of great elongated ridges—like parallel dunes—the highest elevation on them being 2,050 ft., the depressions 1,950 ft. We came to a sweetly pretty streamlet, the Mollah, flowing north into the Paraguay River, and shortly afterwards to the CaittÉ and the Corisho (elev. 1,500 ft.). They were the three real and true sources of the Paraguay, within a short distance of the Seven Lakes.

We had marched 50 kil. that day over rough country. My animals were quite exhausted. Yet early next morning we pushed on once more over transverse undulations and across grassy cuvettes, slightly conical, with circular pools of water in the centre and a florid growth of bamboos in the lowest point of the cuvettes. We ascended over more dyke-like obstructions on our way (elev. 1,700 ft.) and descended once more into a vast basin of campos with stunted trees. At its lowest point there was from north-east to south-west a line of magnificent tall trees. The forest was so dense there that when we entered it we were quite in the dark, as if going through a tunnel. There were fine specimens of various kinds of the jua or juaz or jurubeba (solanum), a medicinal plant 5 to 6 ft. high with enormous dentate leaves—shaped not unlike a vine leaf—possessing upright spikes on their dorsal or mid-rib and on the veins of the leaf.

Then there was plentiful "cepa de pappo," a common liana like a huge boa-constrictor winding its way in a spiral up the tallest trees. I saw some of those liane 3 in. in diameter, with a smooth whitish bark.

The soil at the bottom of the valley (1,500 ft. above sea level) was mostly composed of cinders, but up the slopes white sand was predominant, mixed with ashes. We travelled over a lava flow which formed the bed of the River Macucu, flowing eastward. Guided by the noise, we found a most beautiful waterfall, 100 ft. high, over an extinct circular crater with vertical walls. We kept on rising over a gentle incline, and having reached an elevation of 1,750 ft. we found ourselves suddenly on the upper edge of a great crescent-shaped depression extending in a semicircle from north-east to south-west. Its walls were one-tiered to the west, with a flat table-land on their summit, but were divided into two terraces in the northern part where ranges of hills rose on the plateau.

We had a rapid, steep descent among great rectangular blocks of conglomerate (white marble pebbles embedded in iron rock), great sheets of lava, and sediments of red earth, solidified in places into half-formed rock. I noticed extensive lava flows which had run towards the west; then we came upon extraordinary quantities of loose white marble pebbles and chips. We made our way down upon a kind of spur of red lava, frightfully slippery for my animals. The poor beasts were quite worn out with fatigue.

From the round dome of the headland we perceived to the south a second great circle of flat-topped heights. The immense flow of red lava on which we were radiated terrific heat which it had absorbed from the sun's rays. My dogs, being nearer the ground than we were, had great difficulty in breathing. Their heads and tails hung low, and their tongues dangled fully out of their mouths. They stumbled along panting pitifully. Even we on our mounts felt nearly suffocated by the stifling heat from the sun above and the lava below. The dogs were amusing enough, curling down quickly to rest wherever a mangy shrub gave the slightest suspicion of a shade. The men, more stupid always than beasts, were sweating and swearing freely, and thumped mercilessly on the rumps of the tired animals with the butts and muzzles of their rifles in order to urge them along.

The very sound of the mules' neck-bells seemed tired and worn; its brisk tinkling of our days of vigour had given room to a monotonous and feeble, almost dead, ding ... dong, at long intervals—well suggesting the exhaustion of the poor animals, which were just able to drag along. The slightest obstacle—a loose stone, a step in the lava, and now one animal, then another, would collapse and roll down, and we had to dismount and help them up on their feet again—quite a hard job, I can tell you, when the animals were nearly dead and would not get up again.

As we went along more and more headlands of the great plateau appeared before us to the west. We still went on descending on the top of the long spur of lava. When not too busy with our animals—and quite out of breath with the heat and stifling air from the heated rock—I sometimes glanced at the glorious panorama on both sides of us. When we had proceeded farther I ascertained that there were really two crescents contained side by side within a larger crescent. Under us to the south a vast undulating plain stretched as far as the eye could see towards the south-west and west. On describing a revolution upon your heels your eye met the other end of the larger crescent plateau to the north-west. The Serra do Tombador extended in a south-westerly direction from north of Diamantino to S. Luiz de Caceres, to the west of the Paraguay River. The height of the spur on which we were was 1,350 ft. above the sea level.

We had come in a great circle on the upper edge. A trail could be seen crossing the great undulating valley below us. It passed at the western terminus of the spur we were on. Evidently that was the trail connecting Diamantino with CuyabÁ (the capital of Matto Grosso) via Rosario. The sight of a trail was most exhilarating to my men. Suddenly and quite unexpectedly we came upon a few wretched, tumble-down houses—if one may call them so—smothered in vegetation which grew everywhere. My animals themselves seemed astonished at the unusual sight. The horses neighed and the mules brayed loudly. Masonry work perhaps suggested to them more substantial meals. Down a precipitous ravine, over large boulders and stumbling into big holes, into which the mules disappeared for a few seconds at a time ... there was the main street of Diamantino.

The village—the local people called it "a city"—was the very picture of misery, yet to us it seemed as if we had dropped into the middle of London or Paris. There were a few resident traders, two or three Brazilians, two Italians, and a Turk. All were most hospitable and kind. The chief industry of the place was rubber, which found its way to the coast via the Paraguay River.

Formerly Diamantino was a flourishing place because diamonds were found in abundance. Even now they can be found along the river, but the difficulty of access, even by the easiest way, and the great expense of living there have gradually depopulated the place, which was quite in an abandoned state when I was there.

Here are some of the minimum prices which the rubber collectors had to pay for articles of necessity: Beans, 1s. 6d. to 2s. per litre,[1] or about 4s. a pound; rice, 2s. per litre; flour, 1s. 4d. per litre, about 4s. a pound; sugar, 5s. per kilo (2 pounds), rapadura, or sugar block, 4s. per small cake; tobacco, 5s. per metre of twist; salt, 2s. 8d. to 3s. per litre; coffee, 6s. 6d. per kilo; lard, 6s. 6d. per kilo; purified lard in tins, 16s. to 20s. per 2 kilos. Bars of the commonest laundry soap, 4s. each bar; chickens 10s. to 15s. each; eggs, 10s. to 12s. a dozen; small tins or sardines (containing five sardines) of the most inferior kind, 10s. to 15s. a tin; a one-pound tin of the commonest French salt butter, 15s.

A genial banquet was offered me on my arrival. The school-mistress was set to prepare an excellent and plentiful meal. The mayor and all the notabilities of the place in their Sunday clothing came to fetch me at the house of the firm of Orlando Bros., where I had been most hospitably sheltered, and where I had been requested to wait for them. At the appointed time they arrived—in frock-coats, and each carrying an umbrella.

"Is it raining?" I inquired in my astonishment at seeing the array of articles which I had not seen for several months—especially as a few minutes before I had been outside and it was a lovely starlit night.

"Oh no, indeed, it is not raining; we carry the umbrellas in due honour to you!" they replied in a chorus, accompanied by a grand bow.

This was such an extraordinary compliment that it really took me some time before I could grasp the meaning of it. It seemed that according to the social rules of Diamantino, Matto Grosso, no one could be considered fully dressed unless carrying an umbrella. Rain or shine, the people of Diamantino carried their umbrellas on grand occasions. After that one of the gentlemen pulled out of his pocket a long slip of paper and proceeded to read a speech of welcome. I answered in a few humble words. Another gentleman—there were eight altogether—produced another slip which he duly read in a sonorous voice. Again I replied as best I could. Then, as I was getting really anxious lest some one else should be speechifying again, the mayor of the place offered me his arm, and followed in a most respectful manner by the others, we adjourned to the schoolroom, where the feast was spread upon the table.

More speeches when we entered the room, more speeches before we sat down, speeches in the middle of dinner, speeches after dinner. Unaware of what was coming, I had exhausted all the compliments I could think of in my first speech, and I had to tax my poor brain considerably to reply with grace—especially as I had to speak in Portuguese—to the many charming things which my thoughtful hosts said. The banquet went off well. It is difficult to imagine more considerate, kindly people than those exiles in that far-away spot.

I took careful and repeated astronomical observations for latitude and longitude in order to establish the exact position of that settlement. Lat. 14° 21'·7 S.; Long. 56° 56' W. I purchased all the food I could possibly collect—enough to last us some six months, which cost me a small fortune—as I intended to push out of the place and proceed northward at once.

Four of my men became badly intoxicated upon our arrival. There was another mutiny. They again claimed their pay up to date and wished to leave me. At once they received their money. It was such a relief to me when they went off, even for a few hours, that I was always glad to give them the money and have a short mental rest while they kept away. Unfortunately it was impossible to obtain a single extra man in Diamantino. Labour was scarce, and the few labourers in existence were in absolute slavery. Indeed, slavery existed—it exists to-day—in all Central Brazil, just as it did before slavery was abolished. Only in the old days of legal slavery it was limited to negroes; now the slaves are negroes, mulattoes, white people, even some Europeans. I have seen with my own eyes a German gentleman of refinement in that humble condition.

In the present condition of things the slave, in the first instance, sells himself or is sold by his family. There were indeed few, if any, of the labouring classes in Matto Grosso and Goyaz provinces who were free men or women. All were owned by somebody, and if you wished to employ them—especially to take them away from a village or a city—you had to purchase them from their owners. That meant that if you intended to employ a man—even for a few days—you had to disburse a purchase sum equivalent to two or three hundred pounds sterling, sometimes more. In the following way it was made impossible for the slaves to become free again. Taking advantage of the poverty and vanity of those people, loans of money were offered them in the first instance, and also luxuries in the way of tinned food, clothing, revolvers and rifles. When once they had accepted, and could not repay the sum or value of the articles received, they became the property of the lender, who took good care to increase the debt constantly by supplying cheap articles to them at fifty times their actual cost. The seringueiro, or rubber collector, had a caderneta, or booklet and the master a livro maestro, or account book, in which often double the quantity of articles actually received by the rubber collector were entered. The debt thus increased by leaps and bounds, and in a short time a labourer owed his master, two, three hundred pounds. The rubber collectors tried hard to repay the debt in rubber, which they sold to their masters at a low rate; but it was always easy for the masters to keep the men in debt.

It must be said for the masters that their slaves were not in any way ill-treated; on the contrary—except that a man was seldom given the slightest chance of redeeming himself—they were indeed treated as well as circumstances permitted. Labour, it must be remembered, was so scarce and valuable—it was almost an impossibility to obtain labour in Central Brazil—that it was the care of the master not to lose a labourer.

Much is to be said for the honour of even the worst types of Brazilians. Although many of them would not think twice of murdering or robbing a stranger of all he possessed, they were seldom known to defraud their owners by escaping. A man who ran away from his owner was looked down upon by the entire community. Again, it must be stated that the chances of escape, in those distant regions, were indeed very remote. An escaped slave with no money could not go very far and he would soon die of starvation.

I must confess that, although I tried hard to discover a way by which labour could be obtained and retained in Brazil with the existing laws, I could not find one practicable except that used by the Brazilians, viz. slavery.

The people of Diamantino tried hard to induce one or two men to accompany me—and I was willing to buy them out and eventually would have set them free altogether at the end of the expedition—but they were all so terrified of the Indians if they left the "city" that they preferred to remain slaves.

Alcides had gone round to look for a barber. There was only one in Diamantino, and he was in prison for the murder of his wife, or for some other such trifling matter. Armed with a pair of my scissors, Alcides went to the prison to have his hair cut. Once there he took the opportunity to explain to the prisoner that it could be arranged to procure his escape if he were willing to join the expedition. The barber—who had not inquired which way we should be travelling—jumped at the idea. This necessitated having my hair cut too—rather a trial with scissors that did not cut—in order to arrange matters further in detail. With a special permission from the local authorities the barber was let out accompanied by two policemen—the only two in the place—in order that he might reduce my hair by half its length or more.

While I underwent actual torture in having my hair clipped—as the prisoner's hands were trembling with excitement, and my ears had various narrow escapes—Alcides, who, when he wished, had very persuasive manners, induced not only the prisoner, but the two policemen—all three—to escape and join the expedition. I must say that I did not at all look forward to the prospect of my three new companions; but we were in terrible want of hands. I had visions that my expedition would be entirely wrecked. There was a limit to human endurance and we could not perform miracles. We still had thousands of kilometres to travel over most difficult and dangerous country. Besides, I reflected, after all, I might only be performing an act of kindness by relieving the town of the expense and trouble of keeping its only prisoner, not to speak of the police force.

All was satisfactorily arranged, when the prisoner inquired where we were going. You should have seen his face when I told him.

"No, no, no!" he quickly replied. "No, no, no, no!" and he waved my scissors in the air. "I will not come! I will remain in prison all my life rather than be eaten up by cannibals! No, no, no, no ... no, no, no, no...!" he went on muttering at intervals as he gave the last clipping touches to my hair. He hastened through his job, received his pay in silence, and asked the policemen to take him back quickly to the prison. When the chains, which had temporarily been removed, were put again around his wrists, he departed shaking his head and muttering again—"No, no, no, no...!"

The wise policemen, too, said that naturally, as their prisoner would not escape, they were obliged to remain and keep guard over him ... it was not through lack of courage that they would not come; it was because of their duty!

Of course, Alcides was sadly disappointed, but I was delighted, when it all fell through.

I owe the success of my expeditions to the fact that, no matter what happens, I never will stop anywhere. It is quite fatal, on expeditions of that kind, to stop for any length of time. If you do, the fatigue, the worry, and illness make it generally impossible to start again—all things which you do not feel quite so much as long as you can keep moving. Many a disaster in exploring expeditions could easily have been avoided, had the people known this secret of successful travelling. Push on at all costs—until, of course, you are actually dead.

With my reduced party of two men (Alcides and Filippe) I had to arrange matters differently, and decided to abandon part of my baggage—all things, in fact, which were not absolutely necessary, taking only food, instruments for scientific observations, cameras and photographic plates.

Alcides and Filippe—who by then had become most adventurous—and I were about to start on July 1st, and were making things ready, when two of my deserters returned and begged me to take them along again. They had found living at their own cost rather expensive, and had realized that it would have been an impossibility for them to get out of that place again with the funds at their disposal. Each meal had cost them a small fortune. Animals were extremely expensive, and it was then the wrong season for launches to come up the river as far as Rosario, the nearest port to the south.

"We will come with you," said they, in a sudden outburst of devotion. "We will come. We are brave men. You have always been good and generous to us. We are sorry for what we have done. Order us and we will kill anybody you like for you!" Brazilians of that class have only one idea in their heads—killing, killing, killing!

That was more devotion than I demanded. In order to spare Alcides and Filippe, and myself—as the work thrown upon us would have indeed been beyond our possible strength—I re-employed the two men on the express condition that they should murder no one while they were with me.

At noon of July 1st, accompanied by a mounted escort of honour of the leading citizens with the Mayor at their head, I left Diamantino (elev. 1,030 ft.), travelling north-east. We ascended to the summit of a table-land—the first terrace of which was at an elevation of 1,250 ft., the higher at 1,600 ft. The last words I had heard from a venerable old man as I rode out of Diamantino still rang in my ears.

"You are going to sure death—good-bye!..." On reaching the top of the plateau the courteous friends who had accompanied me also bade me an affectionate farewell. I could see by their faces and their manner that they were saying good-bye to one they believed a doomed man.

"If by chance you come out alive," said the Mayor, in a tentative way, "we should like to have news of you."

On dismal occasions of that kind the sky is always gloomy and black and there is always drizzling rain. So that day, too, the weather did not fail to add to our depressed spirits.

On leaving our friends we started to plunge once more into the unknown. On reaching the top edge of the plateau we witnessed a wonderful sight, rendered more poetic by the slight vagueness of a veil of mist. To the south of Diamantino was the Serra Tombador, extending as far as S. Luiz de Caceres, about 250 kil. as the crow flies to the south-west. Then below us was the LagÕa dos Veados with no outlet, and close by the head-waters of the Rio Preto (a tributary of the Arinos). The Serra do Tombador was parallel nearly all along with the River Paraguay.

Owing to departing so late in the day from Diamantino, and the time we had wasted on the way with social compliments, we were only able to go 12 kil. that afternoon. We halted near the shed of a seringueiro (rubber collector), at an elevation of 1,530 ft., close to the ChapesÀ, a streamlet flowing into the Agua Fria (cold water), which in its turn threw itself into the Rio Preto.

It was muggy and warm during the night—min. 65° Fahr.—with swarms of mosquitoes. We were glad to leave the next morning, following a north-westerly course across a wonderfully beautiful meadow with circular groups of trees and a long belt of vegetation along the stream. It was then that I made my first acquaintance in Brazil with the seringueira (Syphonia elastica or Hevea brasiliensis), which was fairly plentiful in that region. As we shall see, that rubber tree, producing the best rubber known, became more and more common as we proceeded north.

In the cuts of rivers, soft red volcanic rock was exposed, with a surface layer of white sand and grey ashes in the flat meadow. The padding of earth was thin. Except close to rivers and in extinct craters where the accumulations of earth and cinders were often deeper with a good supply of moisture from underneath, the trees were feeble and anÆmic. There again I was amazed to find how unstable and weak most trees were. One could knock them down with a mere hard push—as the roots had no hold in the ground, where they spread horizontally almost on the surface, owing to the rock underneath which prevented their penetrating farther than the thin upper layer of earth, sand, and ashes. If you happened to lean against a tree 4 or 5 in. in diameter, it was not uncommon to see the tree tumble down and you too. The wood also of those trees was very brittle and watery, with no power of resistance worth mentioning.

Many were the streamlets which flowed into the Rio Preto at elevations from 1,450 to 1,500 ft., viz. the Burity Comprido, the Bujui, the Grinko, the Pomba, the Corgo do Campo, the RiberÃo Grande, and the Stiva. Many of those streamlets had beautiful beds of white marble pebbles, which made their cool and clear water look and taste perfectly delicious. Others, with soft black mud bottoms—especially in cuvettes—were extremely troublesome to cross.

On the banks of those streams were marvellous pacobeira palms—a kind of giant banana palm, attaining a height of 30 to 40 ft., with a stem, ovoid in section, of great length, and from which shot out paddle-like leaves of immense size and of a gorgeous green, 6 to 7 ft. long and 3 ft. wide.

On July 3rd we went through thick, dirty, low scrub and forest, except along streams, the banks of which were lined with tall anÆmic trees 1 inch in diameter with a mere bunch of leaves from branches at the summit. We again met with several cuvettes—very grassy, with the usual florid growth of trees in the centre. Those depressions were 1,400 ft. above the sea level. From many of the trees hung huge globes, like tumours. They were nests of cupim, the destructive white ants (termes album), of which there were swarms everywhere in that region. In one night they ate up the bottoms of most of my wooden boxes and rendered many of our possessions useless. They ate up our clothes, injured our saddles by eating the stitching—anything that was not of metal, glass, or polished leather was destroyed by those little devils.

We were beginning to descend gradually on the northern side of the table-land. After crossing a pass 1,350 ft. above the sea level we arrived on a lagoon to our left. Shortly after we reached the left bank of the Arinos River, separated there from the lagoon by a narrow tongue of high land—some 30 ft. high—between the two waters.

It was thus that on July 4th we encamped on that great tributary of the Amazon. We were still thousands of kilometres away from its mouth. My animals were quite exhausted and were unable to continue. Moreover, the forest near this great river—already, so near its birthplace, over 100 metres wide—would have made their coming along quite impossible, as the grazing was getting scarce, and would be scarcer still as we went on north. Then as the River Arinos took me in the direction in which I intended to travel, I had made up my mind to abandon the animals at that spot and attempt to navigate the river—diabolical as its reputation was. We had now travelled on horseback some 2,000 kil. from the last railway station, of which about 600 kil. were over absolutely unknown country. Rough as the travelling had been, it was mere child's play compared with the experiences we had to endure from that day on.

Map showing Author's Route.

Map showing Author's Route.

Map showing the Arinos and Arinos-Juruena Rivers.

Map showing the Arinos and Arinos-Juruena Rivers.

[1] A litre is a cube the sides of which are 3? in.