Along this part of our march there was much of what at a hasty glance seemed to be volcanic rock; but Oliveira showed me that it was a kind of conglomerate, with bubbles or hollows in it, made of sand and iron- bearing earth. He said it was a superficial quaternary deposit formed by erosion from the cretaceous rocks, and that there were here no tertiary deposits. He described the geological structure of the lands through which we had passed as follows: The pantanals were of Pleistocene age. Along the upper Sepotuba, in the region of the rapids, there were sandstones, shales, and clays of Permian age. The rolling country east of this contained eruptive rocks—a porphyritic disbase, with zeolite, quartz, and agate of Triassic age. With the chapadao of the Parecis plateau we came to a land of sand and clay, dotted with lumps of sandstone and pieces of petrified wood; this, according to Oliveira, is of Mesozoic age, possibly cretaceous and similar to the South African formation. There are geologists who consider it as of Permian age.

At Vilhena we were on a watershed which drained into the Gy-Parana, which itself runs into the Madeira nearly midway between its sources and its mouth. A little farther along and northward we again came to streams running ultimately into the Tapajos; and between them, and close to them, were streamlets which drained into the Duvida and Ananas, whose courses and outlets were unknown. This point is part of the divide between the basins of the Madeira and Tapajos. A singular topographical feature of the Plan Alto, the great interior sandy plateau of Brazil, is that at its westernmost end the southward flowing streams, instead of running into the Paraguay as they do farther east, form the headwaters of the Guapore, which may, perhaps, be called the upper main stream of the Madeira. These westernmost streams from the southern edge of the plateau, therefore, begin by flowing south; then for a long stretch they flow southwest; then north, and finally northeast into the Amazon. According to some exceptionally good geological observers, this is probably due to the fact that in a remote geologic past the ocean sent in an arm from the south, between the Plan Alto and what is now the Andean chain. These rivers then emptied into the Andean Sea. The gradual upheaval of the soil has resulted in substituting dry land for this arm of the ocean and in reversing the course of what is now the Madeira, just as, according to these geologists, in somewhat familiar fashion the Amazon has been reversed, it having once been, at least for the upper two thirds of its course, an affluent of the Andean Sea.

From Vilhena we travelled in a generally northward direction. For a few leagues we went across the chapadao, the sands or clays of the nearly level upland plateau, grassy or covered with thin, stunted forest, the same type of country that had been predominant ever since we ascended the Parecis table-land on the morning of the third day after leaving the Sepotuba. Then, at about the point where the trail dipped into a basin containing the head-springs of the Ananas, we left this type of country and began to march through thick forest, not very high. There was little feed for the animals on the Chapadao. There was less in the forest. Moreover, the continual heavy rains made the travelling difficult and laborious for them, and they weakened. However, a couple of marches before we reached Tres Burity, where there is a big ranch with hundreds of cattle, we were met by ten fresh pack-oxen, and our serious difficulties were over.

There were piums in plenty by day, but neither mosquitoes nor sand-flies by night; and for us the trip was very pleasant, save for moments of anxiety about the mules. The loose bullocks furnished us abundance of fresh beef, although, as was inevitable under the circumstances, of a decidedly tough quality. One of the biggest of the bullocks was attacked one night by a vampire bat, and next morning his withers were literally bathed in blood.

With the chapadao we said good-by to the curious, gregarious, and crepuscular or nocturnal spiders which we found so abundant along the line of the telegraph wire. They have offered one of the small problems with which the commission has had to deal. They are not common in the dry season. They swarm during the rains; and, when their tough webs are wet, those that lead from the wire to the ground sometimes effectually short circuit the wire. They have on various occasions caused a good deal of trouble in this manner.

The third night out from Vilhena we emerged for a moment from the endless close-growing forest in which our poor animals got such scanty pickings, and came to a beautiful open country, where grassy slopes, dotted with occasional trees, came down on either side of a little brook which was one of the headwaters of the Duvida. It was a pleasure to see the mules greedily bury their muzzles in the pasturage. Our tents were pitched in the open, near a shady tree, which sent out its low branches on every side. At this camp Cherrie shot a lark, very characteristic of the open upland country, and Miller found two bats in the rotten wood of a dead log. He heard them squeaking and dug them out; he could not tell by what method they had gotten in.

Here Kermit, while a couple of miles from our tents, came across an encampment of Nhambiquaras. There were twenty or thirty of them—men, women, and a few children. Kermit, after the manner of honest folk in the wilderness, advanced ostentatiously in the open, calling out to give warning of his coming. Like surroundings may cause like manners. The early Saxons in England deemed it legal to kill any man who came through the woods without shouting or blowing a horn; and in Nhambiquara land at the present time it is against etiquette, and may be very unhealthy, to come through the woods toward strangers without loudly announcing one's presence. The Nhambiquaras received Kermit with the utmost cordiality, and gave him pineapple-wine to drink. They were stark naked as usual; they had no hammocks or blankets, and their huts were flimsy shelters of palm-branches. Yet they were in fine condition. Half a dozen of the men and a couple of boys accompanied Kermit back to our camp, paying not slightest heed to the rain which was falling. They were bold and friendly, good-natured—at least superficially—and very inquisitive. In feasting, the long reeds thrust through holes in their lips did not seem to bother them, and they laughed at the suggestion of removing them; evidently to have done so would have been rather bad manners—like using a knife as an aid in eating ice-cream. They held two or three dances, and we were again struck by the rhythm and weird, haunting melody of their chanting. After supper they danced beside the camp-fire; and finally, to their delight, most of the members of our own party, Americans and Brazilians, enthusiastically joined the dance, while the colonel and I furnished an appreciative and applauding audience. Next morning, when we were awakened by the chattering and screaming of the numerous macaws, parrots, and parakeets, we found that nearly all the Indians, men and women, were gathered outside the tent. As far as clothing was concerned, they were in the condition of Adam and Eve before the fall. One of the women carried a little squirrel monkey. She put it up the big tree some distance from the tents; and when she called, it came scampering to her across the grass, ran up her, and clung to her neck. They would have liked to pilfer; but as they had no clothes it was difficult for them to conceal anything. One of the women was observed to take a fork; but as she did not possess a rag of clothing of any kind all she did do was to try to bury the fork in the sand and then sit on it; and it was reclaimed without difficulty. One or two of the children wore necklaces and bracelets made of the polished wood of the tucum palm, and of the molars of small rodents.

Next day's march led us across a hilly country of good pastureland. The valleys were densely wooded, palms of several kinds being conspicuous among the other trees; and the brooks at the bottoms we crossed at fords or by the usual rude pole bridges. On the open pastures were occasional trees, usually slender bacaba palms, with heads which the winds had dishevelled until they looked like mops. It was evidently a fine natural cattle country, and we soon began to see scores, perhaps hundreds, of the cattle belonging to the government ranch at Tres Burity, which we reached in the early afternoon. It is beautifully situated: the view roundabout is lovely, and certainly the land will prove healthy when settlements have been definitely established. Here we revelled in abundance of good fresh milk and eggs; and for dinner we had chicken canja and fat beef roasted on big wooden spits; and we even had watermelons. The latter were from seeds brought down by the American engineers who built the Madeira Marmore Railroad—a work which stands honorably distinguished among the many great and useful works done in the development of the tropics of recent years.

Amilcar's pack-oxen, which were nearly worn out, had been left in these fertile pastures. Most of the fresh oxen which he took in their places were unbroken, and there was a perfect circus before they were packed and marched off; in every direction, said the gleeful narrators, there were bucking oxen and loads strewed on the ground. This cattle ranch is managed by the colonel's uncle, his mother's brother, a hale old man of seventy, white-haired but as active and vigorous as ever; with a fine, kindly, intelligent face. His name is Miguel Evangalista. He is a native of Matto Grosso, of practically pure Indian blood, and was dressed in the ordinary costume of the Caboclo—hat, shirt, trousers, and no shoes or stockings. Within the last year he had killed three jaguars, which had been living on the mules; as long as they could get mules they did not at this station molest the cattle.

It was with this uncle's father, Colonel Rondon's own grandfather, that Colonel Rondon as an orphan spent the first seven years of his life. His father died before he was born, and his mother when he was only a year old. He lived on his grandfather's cattle-ranch, some fifty miles from Cuyaba. Then he went to live in Cuyaba with a kinsman on his father's side, from whom he took the name of Rondon; his own father's name was DaSilva. He studied in the Cuyaba Government School, and at sixteen was inscribed as one of the instructors. Then he went to Rio, served for a year in the army as an enlisted man in the ranks, and succeeded finally in getting into the military school. After five years as pupil he served three years as professor of mathematics in this school; and then, as a lieutenant of engineers in the Brazilian army, he came back to his home in Matto Grosso and began his life-work of exploring the wilderness.

Next day we journeyed to the telegraph station at Bonofacio, through alternate spells of glaring sunshine and heavy rain. On the way we stopped at an aldea-village of Nhambiquaras. We first met a couple of men going to hunt, with bows and arrows longer than themselves. A rather comely young woman, carrying on her back a wickerwork basket, or creel, supported by a forehead band, and accompanied by a small child, was with them. At the village there were a number of men, women, and children. Although as completely naked as the others we had met, the members of this band were more ornamented with beads, and wore earrings made from the inside of mussel-shells or very big snail- shells. They were more hairy than the ones we had so far met. The women, but not the men, completely remove the hair from their bodies— and look more, instead of less, indecent in consequence. The chief, whose body was painted red with the juice of a fruit, had what could fairly be styled a mustache and imperial; and one old man looked somewhat like a hairy Ainu, or perhaps even more like an Australian black fellow. My companion told me that this probably represented an infusion of negro blood, and possibly of mulatto blood, from runaway slaves of the old days, when some of the Matto Grosso mines were worked by slave labor. They also thought it possible that this infiltration of African negroes might be responsible for the curious shape of the bigger huts, which were utterly unlike their flimsy, ordinary shelters, and bore no resemblance in shape to those of the other Indian tribes of this region; whereas they were not unlike the ordinary beehive huts of the agricultural African negroes. There were in this village several huts or shelters open at the sides, and two of the big huts. These were of closely woven thatch, circular in outline, with a rounded dome, and two doors a couple of feet high opposite each other, and no other opening. There were fifteen or twenty people to each hut. Inside were their implements and utensils, such as wicker baskets (some of them filled with pineapples), gourds, fire-sticks, wooden knives, wooden mortars, and a board for grating mandioc, made of a thick slab of wood inset with sharp points of a harder wood. From the Brazilians one or two of them had obtained blankets, and one a hammock; and they had also obtained knives, which they sorely needed, for they are not even in the stone age. One woman shielded herself from the rain by holding a green palm-branch down her back. Another had on her head what we at first thought to be a monkey-skin head- dress. But it was a little, live, black monkey. It stayed habitually with its head above her forehead, and its arms and legs spread so that it lay moulded to the shape of her head; but both woman and monkey showed some reluctance about having their photographs taken.

Bonofacio consisted of several thatched one-room cabins, connected by a stockade which was extended to form an enclosure behind them. A number of tame parrots and parakeets, of several different species, scrambled over the roofs and entered the houses. In the open pastures near by were the curious, extensive burrows of a gopher rat, which ate the roots of grass, not emerging to eat the grass but pulling it into the burrows by the roots. These burrows bore a close likeness to those of our pocket gophers. Miller found the animals difficult to trap. Finally, by the aid of Colonel Rondon, several Indians, and two or three of our men, he dug one out. From the central shaft several surface galleries radiated, running for many rods about a foot below the surface, with, at intervals of half a dozen yards, mounds where the loose earth had been expelled. The central shaft ran straight down for about eight feet, and then laterally for about fifteen feet, to a kind of chamber. The animal dug hard to escape, but when taken and put on the surface of the ground it moved slowly and awkwardly. It showed vicious courage. In looks it closely resembled our pocket gophers, but it had no pockets. This was one of the most interesting small mammals that we secured.

After breakfast at Bonofacio a number of Nhambiquaras—men, women, and children—strolled in. The men gave us an exhibition of not very good archery; when the bow was bent, it was at first held so that the arrow pointed straight upwards and was then lowered so that the arrow was aimed at the target. Several of the women had been taken from other tribes, after their husbands or fathers had been killed; for the Nhambiquaras are light-hearted robbers and murderers. Two or three miserable dogs accompanied them, half-starved and mangy, but each decorated with a collar of beads. The headmen had three or four wives apiece, and the women were the burden-bearers, but apparently were not badly treated. Most of them were dirty, although well-fed looking, and their features were of a low type; but some, especially among the children, were quite attractive.

From Bonofacio we went about seven miles, across a rolling prairie dotted with trees and clumps of shrub. There, on February 24, we joined Amilcar, who was camped by a brook which flowed into the Duvida. We were only some six miles from our place of embarkation on the Duvida, and we divided our party and our belongings. Amilcar, Miller, Mello, and Oliveira were to march three days to the Gy-Parana, and then descend it, and continue down the Madeira to Manaos. Rondon, Lyra, the doctor, Cherrie, Kermit, and I, with sixteen paddlers, in seven canoes, were to descend the Duvida, and find out whether it led into the Gy-Parana, our purpose was to return and descend the Ananas, whose outlet was also unknown. Having this in view, we left a fortnight's provisions for our party of six at Bonofacio. We took with us provisions for about fifty days; not full rations, for we hoped in part to live on the country—on fish, game, nuts, and palm-tops. Our personal baggage was already well cut down: Cherrie, Kermit, and I took the naturalist's fly to sleep under, and a very light little tent extra for any one who might fall sick. Rondon, Lyra, and the doctor took one of their own tents. The things that we carried were necessities—food, medicines, bedding, instruments for determining the altitude and longitude and latitude—except a few books, each in small compass: Lyra's were in German, consisting of two tiny volumes of Goethe and Schiller; Kermit's were in Portuguese; mine, all in English, included the last two volumes of Gibbon, the plays of Sophocles, More's "Utopia," Marcus Aurelius, and Epictetus, the two latter lent me by a friend, Major Shipton of the regulars, our military attache at Buenos Aires.

If our canoe voyage was prosperous we would gradually lighten the loads by eating the provisions. If we met with accidents, such as losing canoes and men in the rapids, or losing men in encounters with Indians, or if we encountered overmuch fever and dysentery, the loads would lighten themselves. We were all armed. We took no cartridges for sport. Cherrie had some to be used sparingly for collecting specimens. The others were to be used—unless in the unlikely event of having to repel an attack—only to procure food. The food and the arms we carried represented all reasonable precautions against suffering and starvation; but, of course, if the course of the river proved very long and difficult, if we lost our boats over falls or in rapids, or had to make too many and too long portages, or were brought to a halt by impassable swamps, then we would have to reckon with starvation as a possibility. Anything might happen. We were about to go into the unknown, and no one could say what it held.

NOTE: The first four days, before we struck the upper rapids, and during which we made nearly seventy kilometres, are of course not included when I speak of our making our way down the rapids.

I hope that this year the Ananas, or Pineapple, will also be put on the map. One of Colonel Rondon's subordinates is to attempt the descent of the river. We passed the headwaters of the Pineapple on the high plateau, very possibly we passed its mouth, although it is also possible that it empties into the Canama or Tapajos. But it will not be "put on the map" until some one descends and finds out where, as a matter of fact, it really does go.

It would be well if a geographical society of standing would investigate the formal and official charges made by Colonel Rondon, an officer and gentleman of the highest repute, against Mr. Savage Landor. Colonel Rondon, in an official report to the Brazilian Government, has written a scathing review of Mr. Landor. He states that Mr. Savage Landor did not perform, and did not even attempt to perform, the work he had contracted to do in exploration for the Brazilian Government. Mr. Landor had asserted and promised that he would go through unknown country along the line of eleven degrees latitude south, and, as Colonel Rondon states, it was because of this proposal of his that the Brazilian Government gave him material financial assistance in advance. However, Colonel Rondon sets forth that Mr. Landor did not keep his word or make any serious effort to fulfil his moral obligation to do as he had said he would do. In a letter to me under date of May 1, 1914—a letter which has been published in full in France—Colonel Rondon goes at length into the question of what territory Mr. Landor had traversed. Colonel Rondon states that—excepting on one occasion, when Mr. Landor, wandering off a beaten trail, immediately got lost and shortly returned to his starting-point without making any discoveries—he kept to old, well- travelled routes. One sentence of the colonel's letter to me runs as follows: "I can guarantee to you that in Brazil Mr. Landor did not cross a hand's breadth of land that had not been explored, the greater part of it many centuries ago." As regards Mr. Landor's sole and brief experience in leaving a beaten route, Colonel Rondon states that at Sao Manoel Mr. Landor engaged from Senhor Jose Sotero Barreto (the revenue officer of Matto Grosso, at Sao Manoel) a guide to lead him across a well-travelled trail which connects the Tapajos with the Madeira via the Canama. The guide, however, got lost, and after a few days they all returned to the point of departure instead of going through to the Canama.

Senhor Barreto, a gentleman of high standing, related this last incident to Fiala when Fiala descended the Tapajos (and, by the way, Fiala's trip down the Papagaio, Juruena, and Tapajos was infinitely more important than all the work Mr. Landor did in South America put together). Lieutenants Pyrineus and Mello, mentioned in the body of this work, informed me that they accompanied Mr. Landor on most of his overland trip before he embarked on the Arinos, and that he simply followed the highroad or else the telegraph-line, and furthermore, Colonel Rondon states that the Indians whom Mr. Landor encountered and photographed were those educated at the missions.

Colonel Rondon's official report to the Brazilian Government and his letter to me are of interest to all geographers and other scientific men who have any concern with the alleged discoveries of Mr. Landor. They contain very grave charges, with which it is not necessary for me to deal. Suffice it to say that Mr. Landor's accounts of his alleged exploration cannot be considered as entitled to the slightest serious consideration until he has satisfactorily and in detail answered Colonel Rondon; and this he has thus far signally failed to do.

Fortunately, there are numerous examples of exactly the opposite type of work. From the days of Humboldt and Spix and Martius to the present time, German explorers have borne a conspicuous part in the exploration of South America. As representatives of the men and women who have done such capital work, who have fronted every hazard and hardship and labored in the scientific spirit, and who have added greatly to our fund of geographic, biologic, and ethnographic knowledge, I may mention Miss Snethlage and Herr Karl von den Steinen.



VIII. THE RIVER OF DOUBT

On February 27, 1914, shortly after midday, we started down the River of Doubt into the unknown. We were quite uncertain whether after a week we should find ourselves in the Gy-Parana, or after six weeks in the Madeira, or after three months we knew not where. That was why the river was rightly christened the Duvida.

We had been camped close to the river, where the trail that follows the telegraph line crosses it by a rough bridge. As our laden dugouts swung into the stream, Amilcar and Miller and all the others of the Gy-Parana party were on the banks and the bridge to wave farewell and wish us good-by and good luck. It was the height of the rainy season, and the swollen torrent was swift and brown. Our camp was at about 12 degrees 1 minute latitude south and 60 degrees 15 minutes longitude west of Greenwich. Our general course was to be northward toward the equator, by waterway through the vast forest.

We had seven canoes, all of them dugouts. One was small, one was cranky, and two were old, waterlogged, and leaky. The other three were good. The two old canoes were lashed together, and the cranky one was lashed to one of the others. Kermit with two paddlers went in the smallest of the good canoes; Colonel Rondon and Lyra with three other paddlers in the next largest; and the doctor, Cherrie, and I in the largest with three paddlers. The remaining eight camaradas—there were sixteen in all—were equally divided between our two pairs of lashed canoes. Although our personal baggage was cut down to the limit necessary for health and efficiency, yet on such a trip as ours, where scientific work has to be done and where food for twenty-two men for an unknown period of time has to be carried, it is impossible not to take a good deal of stuff; and the seven dugouts were too heavily laden.

The paddlers were a strapping set. They were expert rivermen and men of the forest, skilled veterans in wilderness work. They were lithe as panthers and brawny as bears. They swam like waterdogs. They were equally at home with pole and paddle, with axe and machete; and one was a good cook and others were good men around camp. They looked like pirates in the pictures of Howard Pyle or Maxfield Parrish; one or two of them were pirates, and one worse than a pirate; but most of them were hard-working, willing, and cheerful. They were white,—or, rather, the olive of southern Europe,—black, copper-colored, and of all intermediate shades. In my canoe Luiz the steersman, the headman, was a Matto Grosso negro; Julio the bowsman was from Bahia and of pure Portuguese blood; and the third man, Antonio, was a Parecis Indian.

The actual surveying of the river was done by Colonel Rondon and Lyra, with Kermit as their assistant. Kermit went first in his little canoe with the sighting-rod, on which two disks, one red and one white, were placed a metre apart. He selected a place which commanded as long vistas as possible up-stream and down, and which therefore might be at the angle of a bend; landed; cut away the branches which obstructed the view; and set up the sighting-pole—incidentally encountering maribundi wasps and swarms of biting and stinging ants. Lyra, from his station up-stream, with his telemetre established the distance, while Colonel Rondon with the compass took the direction, and made the records. Then they moved on to the point Kermit had left, and Kermit established a new point within their sight. The first half-day's work was slow. The general course of the stream was a trifle east of north, but at short intervals it bent and curved literally toward every point of the compass. Kermit landed nearly a hundred times, and we made but nine and a third kilometres.

My canoe ran ahead of the surveying canoes. The height of the water made the going easy, for most of the snags and fallen trees were well beneath the surface. Now and then, however, the swift water hurried us toward ripples that marked ugly spikes of sunken timber, or toward uprooted trees that stretched almost across the stream. Then the muscles stood out on the backs and arms of the paddlers as stroke on stroke they urged us away from and past the obstacle. If the leaning or fallen trees were the thorny, slender-stemmed boritana palms, which love the wet, they were often, although plunged beneath the river, in full and vigorous growth, their stems curving upward, and their frond- crowned tops shaken by the rushing water. It was interesting work, for no civilized man, no white man, had ever gone down or up this river or seen the country through which we were passing. The lofty and matted forest rose like a green wall on either hand. The trees were stately and beautiful. The looped and twisted vines hung from them like great ropes. Masses of epiphytes grew both on the dead trees and the living; some had huge leaves like elephants' ears. Now and then fragrant scents were blown to us from flowers on the banks. There were not many birds, and for the most part the forest was silent; rarely we heard strange calls from the depths of the woods, or saw a cormorant or ibis.

My canoe ran only a couple of hours. Then we halted to wait for the others. After a couple of hours more, as the surveyors had not turned up, we landed and made camp at a spot where the bank rose sharply for a hundred yards to a level stretch of ground. Our canoes were moored to trees. The axemen cleared a space for the tents; they were pitched, the baggage was brought up, and fires were kindled. The woods were almost soundless. Through them ran old tapir trails, but there was no fresh sign. Before nightfall the surveyors arrived. There were a few piums and gnats, and a few mosquitoes after dark, but not enough to make us uncomfortable. The small stingless bees, of slightly aromatic odor, swarmed while daylight lasted and crawled over our faces and hands; they were such tame, harmless little things that when they tickled too much I always tried to brush them away without hurting them. But they became a great nuisance after a while. It had been raining at intervals, and the weather was overcast; but after the sun went down the sky cleared. The stars were brilliant overhead, and the new moon hung in the west. It was a pleasant night, the air almost cool, and we slept soundly.

Next morning the two surveying canoes left immediately after breakfast. An hour later the two pairs of lashed canoes pushed off. I kept our canoe to let Cherrie collect, for in the early hours we could hear a number of birds in the woods near by. The most interesting birds he shot were a cotinga, brilliant turquoise-blue with a magenta- purple throat, and a big woodpecker, black above and cinnamon below with an entirely red head and neck. It was almost noon before we started. We saw a few more birds; there were fresh tapir and paca tracks at one point where we landed; once we heard howler monkeys from the depth of the forest, and once we saw a big otter in midstream. As we drifted and paddled down the swirling brown current, through the vivid rain-drenched green of the tropic forest, the trees leaned over the river from both banks. When those that had fallen in the river at some narrow point were very tall, or where it happened that two fell opposite each other, they formed barriers which the men in the leading canoes cleared with their axes. There were many palms, both the burity with its stiff fronds like enormous fans, and a handsome species of bacaba, with very long, gracefully curving fronds. In places the palms stood close together, towering and slender, their stems a stately colonnade, their fronds an arched fretwork against the sky. Butterflies of many hues fluttered over the river. The day was overcast, with showers of rain. When the sun broke through rifts in the clouds, his shafts turned the forest to gold.

In mid-afternoon we came to the mouth of a big and swift affluent entering from the right. It was undoubtedly the Bandeira, which we had crossed well toward its head, some ten days before, on our road to Bonofacio. The Nhambiquaras had then told Colonel Rondon that it flowed into the Duvida. After its junction, with the added volume of water, the river widened without losing its depth. It was so high that it had overflowed and stood among the trees on the lower levels. Only the higher stretches were dry. On the sheer banks where we landed we had to push the canoes for yards or rods through the branches of the submerged trees, hacking and hewing. There were occasional bays and ox-bows from which the current had shifted. In these the coarse marsh grass grew tall.

This evening we made camp on a flat of dry ground, densely wooded, of course, directly on the edge of the river and five feet above it. It was fine to see the speed and sinewy ease with which the choppers cleared an open space for the tents. Next morning, when we bathed before sunrise, we dived into deep water right from the shore, and from the moored canoes. This second day we made sixteen and a half kilometres along the course of the river, and nine kilometres in a straight line almost due north.

The following day, March 1, there was much rain—sometimes showers, sometimes vertical sheets of water. Our course was somewhat west of north and we made twenty and a half kilometres. We passed signs of Indian habitation. There were abandoned palm-leaf shelters on both banks. On the left bank we came to two or three old Indian fields, grown up with coarse fern and studded with the burned skeletons of trees. At the mouth of a brook which entered from the right some sticks stood in the water, marking the site of an old fish-trap. At one point we found the tough vine hand-rail of an Indian bridge running right across the river, a couple of feet above it. Evidently the bridge had been built at low water. Three stout poles had been driven into the stream-bed in a line at right angles to the current. The bridge had consisted of poles fastened to these supports, leading between them and from the support at each end to the banks. The rope of tough vines had been stretched as a hand-rail, necessary with such precarious footing. The rise of the river had swept away the bridge, but the props and the rope hand-rail remained. In the afternoon, from the boat, Cherrie shot a large dark-gray monkey with a prehensile tail. It was very good eating.

We camped on a dry level space, but a few feet above, and close beside, the river—so that our swimming-bath was handy. The trees were cleared and camp was made with orderly hurry. One of the men almost stepped on a poisonous coral-snake, which would have been a serious thing, as his feet were bare. But I had on stout shoes, and the fangs of these serpents—unlike those of the pit-vipers—are too short to penetrate good leather. I promptly put my foot on him, and he bit my shoe with harmless venom. It has been said that the brilliant hues of the coral-snake when in its native haunts really confer on it a concealing coloration. In the dark and tangled woods, and to an only less extent in the ordinary varied landscape, anything motionless, especially if partially hidden, easily eludes the eye. But against the dark-brown mould of the forest floor on which we found this coral- snake its bright and varied coloration was distinctly revealing; infinitely more so than the duller mottling of the jararaca and other dangerous snakes of the genus lachecis. In the same place, however, we found a striking example of genuine protective or mimetic coloration and shape. A rather large insect larva—at least we judged it to be a larval form, but we were none of us entomologists—bore a resemblance to a partially curled dry leaf which was fairly startling. The tail exactly resembled the stem or continuation of the midrib of the dead leaf. The flattened body was curled up at the sides, and veined and colored precisely like the leaf. The head, colored like the leaf, projected in front.

We were still in the Brazilian highlands. The forest did not teem with life. It was generally rather silent; we did not hear such a chorus of birds and mammals as we had occasionally heard even on our overland journey, when more than once we had been awakened at dawn by the howling, screaming, yelping, and chattering of monkeys, toucans, macaws, parrots, and parakeets. There were, however, from time to time, queer sounds from the forest, and after nightfall different kinds of frogs and insects uttered strange cries and calls. In volume and frequency these seemed to increase until midnight. Then they died away and before dawn everything was silent.

At this camp the carregadores ants completely devoured the doctor's undershirt, and ate holes in his mosquito-net; and they also ate the strap of Lyra's gun-case. The little stingless bees, of many kinds, swarmed in such multitudes, and were so persevering, that we had to wear our head-nets when we wrote or skinned specimens.

The following day was almost without rain. It was delightful to drift and paddle slowly down the beautiful tropical river. Until mid- afternoon the current was not very fast, and the broad, deep, placid stream bent and curved in every direction, although the general course was northwest. The country was flat, and more of the land was under than above water. Continually we found ourselves travelling between stretches of marshy forest where for miles the water stood or ran among the trees. Once we passed a hillock. We saw brilliantly colored parakeets and trogons. At last the slow current quickened. Faster it went, and faster, until it began to run like a mill-race, and we heard the roar of rapids ahead. We pulled to the right bank, moored the canoes, and while most of the men pitched camp two or three of them accompanied us to examine the rapids. We had made twenty kilometres.

We soon found that the rapids were a serious obstacle. There were many curls, and one or two regular falls, perhaps six feet high. It would have been impossible to run them, and they stretched for nearly a mile. The carry, however, which led through woods and over rocks in a nearly straight line, was somewhat shorter. It was not an easy portage over which to carry heavy loads and drag heavy dugout canoes. At the point where the descent was steepest there were great naked flats of friable sandstone and conglomerate. Over parts of these, where there was a surface of fine sand, there was a growth of coarse grass. Other parts were bare and had been worn by the weather into fantastic shapes—one projection looked like an old-fashioned beaver hat upside down. In this place, where the naked flats of rock showed the projection of the ledge through which the river had cut its course, the torrent rushed down a deep, sheer-sided, and extremely narrow channel. At one point it was less than two yards across, and for quite a distance not more than five or six yards. Yet only a mile or two above the rapids the deep, placid river was at least a hundred yards wide. It seemed extraordinary, almost impossible, that so broad a river could in so short a space of time contract its dimensions to the width of the strangled channel through which it now poured its entire volume.

This has for long been a station where the Nhambiquaras at intervals built their ephemeral villages and tilled the soil with the rude and destructive cultivation of savages. There were several abandoned old fields, where the dense growth of rank fern hid the tangle of burnt and fallen logs. Nor had the Nhambiquaras been long absent. In one trail we found what gypsies would have called a "pateran," a couple of branches arranged crosswise, eight leaves to a branch; it had some special significance, belonging to that class of signals, each with some peculiar and often complicated meaning, which are commonly used by many wild peoples. The Indians had thrown a simple bridge, consisting of four long poles, without a hand-rail, across one of the narrowest parts of the rock gorge through which the river foamed in its rapid descent. This sub-tribe of Indians was called the Navaite; we named the rapids after them, Navaite Rapids. By observation Lyra found them to be (in close approximation to) latitude 11 degrees 44 minutes south and longitude 60 degrees 18 minutes west from Greenwich.

We spent March 3 and 4 and the morning of the 5th in portaging around the rapids. The first night we camped in the forest beside the spot where we had halted. Next morning we moved the baggage to the foot of the rapids, where we intended to launch the canoes, and pitched our tents on the open sandstone flat. It rained heavily. The little bees were in such swarms as to be a nuisance. Many small stinging bees were with them, which stung badly. We were bitten by huge horse-flies, the size of bumblebees. More serious annoyance was caused by the pium and boroshuda flies during the hours of daylight, and by the polvora, the sand-flies, after dark. There were a few mosquitoes. The boroshudas were the worst pests; they brought the blood at once, and left marks that lasted for weeks. I did my writing in head-net and gauntlets. Fortunately we had with us several bottles of "fly dope"—so named on the label—put up, with the rest of our medicine, by Doctor Alexander Lambert; he had tested it in the north woods and found it excellent. I had never before been forced to use such an ointment, and had been reluctant to take it with me; but now I was glad enough to have it, and we all of us found it exceedingly useful. I would never again go into mosquito or sand-fly country without it. The effect of an application wears off after half an hour or so, and under many conditions, as when one is perspiring freely, it is of no use; but there are times when minute mosquitoes and gnats get through head-nets and under mosquito-bars, and when the ointments occasionally renewed may permit one to get sleep or rest which would otherwise be impossible of attainment. The termites got into our tent on the sand- flat, ate holes in Cherrie's mosquito-net and poncho, and were starting to work at our duffel-bags, when we discovered them.

Packing the loads across was simple. Dragging the heavy dugouts was labor. The biggest of the two water-logged ones was the heaviest. Lyra and Kermit did the job. All the men were employed at it except the cook, and one man who was down with fever. A road was chopped through the forest and a couple of hundred stout six-foot poles, or small logs, were cut as rollers and placed about two yards apart. With block and tackle the seven dugouts were hoisted out of the river up the steep banks, and up the rise of ground until the level was reached. Then the men harnessed themselves two by two on the drag-rope, while one of their number pried behind with a lever, and the canoe, bumping and sliding, was twitched through the woods. Over the sandstone flats there were some ugly ledges, but on the whole the course was down-hill and relatively easy. Looking at the way the work was done, at the good-will, the endurance, and the bull-like strength of the camaradas, and at the intelligence and the unwearied efforts of their commanders, one could but wonder at the ignorance of those who do not realize the energy and the power that are so often possessed by, and that may be so readily developed in, the men of the tropics. Another subject of perpetual wonder is the attitude of certain men who stay at home, and still more the attitude of certain men who travel under easy conditions, and who belittle the achievements of the real explorers of, the real adventures in, the great wilderness. The impostors and romancers among explorers or would-be explorers and wilderness wanderers have been unusually prominent in connection with South America (although the conspicuous ones are not South Americans, by the way); and these are fit subjects for condemnation and derision. But the work of the genuine explorer and wilderness wanderer is fraught with fatigue, hardship, and danger. Many of the men of little knowledge talk glibly of portaging as if it were simple and easy. A portage over rough and unknown ground is always a work of difficulty and of some risk to the canoe; and in the untrodden, or even in the unfrequented, wilderness risk to the canoe is a serious matter. This particular portage at Navaite Rapids was far from being unusually difficult; yet it not only cost two and a half days of severe and incessant labor, but it cost something in damage to the canoes. One in particular, the one in which I had been journeying, was split in a manner which caused us serious uneasiness as to how long, even after being patched, it would last. Where the canoes were launched, the bank was sheer, and one of the water-logged canoes filled and went to the bottom; and there was more work in raising it.

We were still wholly unable to tell where we were going or what lay ahead of us. Round the camp-fire, after supper, we held endless discussions and hazarded all kinds of guesses on both subjects. The river might bend sharply to the west and enter the Gy-Parana high up or low down, or go north to the Madeira, or bend eastward and enter the Tapajos, or fall into the Canuma and finally through one of its mouths enter the Amazon direct. Lyra inclined to the first, and Colonel Rondon to the second, of these propositions. We did not know whether we had one hundred or eight hundred kilometres to go, whether the stream would be fairly smooth or whether we would encounter waterfalls, or rapids, or even some big marsh or lake. We could not tell whether or not we would meet hostile Indians, although no one of us ever went ten yards from camp without his rifle. We had no idea how much time the trip would take. We had entered a land of unknown possibilities.

We started down-stream again early in the afternoon of March 5. Our hands and faces were swollen from the bites and stings of the insect pests at the sand-flat camp, and it was a pleasure once more to be in the middle of the river, where they did not come, in any numbers, while we were in motion. The current was swift, but the river was so deep that there were no serious obstructions. Twice we went down over slight riffles, which in the dry season were doubtless rapids; and once we struck a spot where many whirlpools marked the presence underneath of boulders which would have been above water had not the river been so swollen by the rains. The distance we covered in a day going down-stream would have taken us a week if we had been going up. The course wound hither and thither, sometimes in sigmoid curves; but the general direction was east of north. As usual, it was very beautiful; and we never could tell what might appear around any curve. In the forest that rose on either hand were tall rubber-trees. The surveying canoes, as usual, went first, while I shepherded the two pairs of lashed cargo canoes. I kept them always between me and the surveying canoes—ahead of me until I passed the surveying canoes, then behind me until, after an hour or so, I had chosen a place to camp. There was so much overflowed ground that it took us some little time this afternoon before we found a flat place high enough to be dry. Just before reaching camp Cherrie shot a jacu, a handsome bird somewhat akin to, but much smaller than, a turkey; after Cherrie had taken its skin, its body made an excellent canja. We saw parties of monkeys; and the false bellbirds uttered their ringing whistles in the dense timber around our tents. The giant ants, an inch and a quarter long, were rather too plentiful around this camp; one stung Kermit; it was almost like the sting of a small scorpion, and pained severely for a couple of hours. This half-day we made twelve kilometres.

On the following day we made nineteen kilometres, the river twisting in every direction, but in its general course running a little west of north. Once we stopped at a bee-tree, to get honey. The tree was a towering giant, of the kind called milk-tree, because a thick milky juice runs freely from any cut. Our camaradas eagerly drank the white fluid that flowed from the wounds made by their axes. I tried it. The taste was not unpleasant, but it left a sticky feeling in the mouth. The helmsman of my boat, Luiz, a powerful negro, chopped into the tree, balancing himself with springy ease on a slight scaffolding. The honey was in a hollow, and had been made by medium-sized stingless bees. At the mouth of the hollow they had built a curious entrance of their own, in the shape of a spout of wax about a foot long. At the opening the walls of the spout showed the wax formation, but elsewhere it had become in color and texture indistinguishable from the bark of the tree. The honey was delicious, sweet and yet with a tart flavor. The comb differed much from that of our honey-bees. The honey-cells were very large, and the brood-cells, which were small, were in a single instead of a double row. By this tree I came across an example of genuine concealing coloration. A huge tree-toad, the size of a bullfrog, was seated upright—not squatted flat—on a big rotten limb. It was absolutely motionless; the yellow brown of its back, and its dark sides, exactly harmonized in color with the light and dark patches on the log; the color was as concealing, here in its natural surroundings, as is the color of our common wood-frog among the dead leaves of our woods. When I stirred it up it jumped to a small twig, catching hold with the disks of its finger-tips, and balancing itself with unexpected ease for so big a creature, and then hopped to the ground and again stood motionless. Evidently it trusted for safety to escaping observation. We saw some monkeys and fresh tapir sign, and Kermit shot a jacu for the pot.

At about three o'clock I was in the lead, when the current began to run more quickly. We passed over one or two decided ripples, and then heard the roar of rapids ahead, while the stream began to race. We drove the canoe into the bank, and then went down a tapir trail, which led alongside the river, to reconnoiter. A quarter of a mile's walk showed us that there were big rapids, down which the canoes could not go; and we returned to the landing. All the canoes had gathered there, and Rondon, Lyra, and Kermit started down-stream to explore. They returned in an hour, with the information that the rapids continued for a long distance, with falls and steep pitches of broken water, and that the portage would take several days. We made camp just above the rapids. Ants swarmed, and some of them bit savagely. Our men, in clearing away the forest for our tents, left several very tall and slender accashy palms; the bole of this palm is as straight as an arrow and is crowned with delicate, gracefully curved fronds. We had come along the course of the river almost exactly a hundred kilometres; it had twisted so that we were only about fifty-five kilometres north of our starting-point. The rock was porphyritic.

The 7th, 8th, and 9th we spent in carrying the loads and dragging and floating the dugouts past the series of rapids at whose head we had stopped.

The first day we shifted camp a kilometre and a half to the foot of this series of rapids. This was a charming and picturesque camp. It was at the edge of the river, where there was a little, shallow bay with a beach of firm sand. In the water, at the middle point of the beach, stood a group of three burity palms, their great trunks rising like columns. Round the clearing in which our tents stood were several very big trees; two of them were rubber-trees. Kermit went down-stream five or six kilometres, and returned, having shot a jacu and found that at the point which he had reached there was another rapids, almost a fall, which would necessitate our again dragging the canoes over a portage. Antonio, the Parecis, shot a big monkey; of this I was glad because portaging is hard work, and the men appreciated the meat. So far Cherrie had collected sixty birds on the Duvida, all of them new to the collection, and some probably new to science. We saw the fresh sign of paca, agouti, and the small peccary, and Kermit with the dogs roused a tapir, which crossed the river right through the rapids; but no one got a shot at it.

Except at one or perhaps two points a very big dugout, lightly loaded, could probably run all these rapids. But even in such a canoe it would be silly to make the attempt on an exploring expedition, where the loss of a canoe or of its contents means disaster; and moreover such a canoe could not be taken, for it would be impossible to drag it over the portages on the occasions when the portages became inevitable. Our canoes would not have lived half a minute in the wild water.

On the second day the canoes and loads were brought down to the foot of the first rapids. Lyra cleared the path and laid the logs for rollers, while Kermit dragged the dugouts up the bank from the water with block and tackle, with strain of rope and muscle. Then they joined forces, as over the uneven ground it needed the united strength of all their men to get the heavy dugouts along. Meanwhile the colonel with one attendant measured the distance, and then went on a long hunt, but saw no game. I strolled down beside the river for a couple of miles, but also saw nothing. In the dense tropical forest of the Amazonian basin hunting is very difficult, especially for men who are trying to pass through the country as rapidly as possible. On such a trip as ours getting game is largely a matter of chance.

On the following day Lyra and Kermit brought down the canoes and loads, with hard labor, to the little beach by the three palms where our tents were pitched. Many pacovas grew round about. The men used their immense leaves, some of which were twelve feet long and two and a half feet broad, to roof the flimsy shelters under which they hung their hammocks. I went into the woods, but in the tangle of vegetation it would have been a mere hazard had I seen any big animal. Generally the woods were silent and empty. Now and then little troops of birds of many kinds passed—wood-hewers, ant-thrushes, tanagers, flycatchers; as in the spring and fall similar troops of warblers, chickadees, and nuthatches pass through our northern woods. On the rocks and on the great trees by the river grew beautiful white and lilac orchids, the sobralia, of sweet and delicate fragrance. For the moment my own books seemed a trifle heavy, and perhaps I would have found the day tedious if Kermit had not lent me the Oxford Book of French Verse. Eustache Deschamp, Joachim du Bellay, Ronsard, the delightful La Fontaine, the delightful but appalling Villon, Victor Hugo's "Guitare," Madame Desbordes-Valmore's lines on the little girl and her pillow, as dear little verses about a child as ever were written—these and many others comforted me much, as I read them in head-net and gauntlets, sitting on a log by an unknown river in the Amazonian forest.

On the 10th we again embarked and made a kilometre and a half, spending most of the time in getting past two more rapids. Near the first of these we saw a small cayman, a jacare-tinga. At each set of rapids the canoes were unloaded and the loads borne past on the shoulders of the camaradas; three of the canoes were paddled down by a couple of naked paddlers apiece; and the two sets of double canoes were let down by ropes, one of one couple being swamped but rescued and brought safely to shore on each occasion. One of the men was upset while working in the swift water, and his face was cut against the stones. Lyra and Kermit did the actual work with the camaradas. Kermit, dressed substantially like the camaradas themselves, worked in the water, and, as the overhanging branches were thronged with crowds of biting and stinging ants, he was marked and blistered over his whole body. Indeed, we all suffered more or less from these ants; while the swarms of biting flies grew constantly more numerous. The termites ate holes in my helmet and also in the cover of my cot. Every one else had a hammock. At this camp we had come down the river about 102 kilometres, according to the surveying records, and in height had descended nearly 100 metres, as shown by the aneroid—although the figure in this case is only an approximation, as an aneroid cannot be depended on for absolute accuracy of results.

Next morning we found that during the night we had met with a serious misfortune. We had halted at the foot of the rapids. The canoes were moored to trees on the bank, at the tail of the broken water. The two old canoes, although one of them was our biggest cargo-carrier, were water-logged and heavy, and one of them was leaking. In the night the river rose. The leaky canoe, which at best was too low in the water, must have gradually filled from the wash of the waves. It sank, dragging down the other; they began to roll, bursting their moorings; and in the morning they had disappeared. A canoe was launched to look for them; but, rolling over the boulders on the rocky bottom, they had at once been riven asunder, and the big fragments that were soon found, floating in eddies, or along the shore, showed that it was useless to look farther. We called these rapids Broken Canoe Rapids.

It was not pleasant to have to stop for some days; thanks to the rapids, we had made slow progress, and with our necessarily limited supply of food, and no knowledge whatever of what was ahead of us, it was important to make good time. But there was no alternative. We had to build either one big canoe or two small ones. It was raining heavily as the men started to explore in different directions for good canoe trees. Three—which ultimately proved not very good for the purpose—were found close to camp; splendid-looking trees, one of them five feet in diameter three feet from the ground. The axemen immediately attacked this one under the superintendence of Colonel Rondon. Lyra and Kermit started in opposite directions to hunt. Lyra killed a jacu for us, and Kermit killed two monkeys for the men. Toward night fall it cleared. The moon was nearly full, and the foaming river gleamed like silver.

Our men were "regional volunteers," that is, they had enlisted in the service of the Telegraphic Commission especially to do this wilderness work, and were highly paid, as was fitting, in view of the toil, hardship, and hazard to life and health. Two of them had been with Colonel Rondon during his eight months' exploration in 1909, at which time his men were regulars, from his own battalion of engineers. His four aides during the closing months of this trip were Lieutenants Lyra, Amarante, Alencarliense, and Pyrineus. The naturalist Miranda Ribeiro also accompanied him. This was the year when, marching on foot through an absolutely unknown wilderness, the colonel and his party finally reached the Gy-Parana, which on the maps was then (and on most maps is now) placed in an utterly wrong course, and over a degree out of its real position. When they reached the affluents of the Gy-Parana a third of the members of the party were so weak with fever that they could hardly crawl. They had no baggage. Their clothes were in tatters, and some of the men were almost naked. For months they had had no food except what little game they shot, and especially the wild fruits and nuts; if it had not been for the great abundance of the Brazil-nuts they would all have died. At the first big stream they encountered they built a canoe, and Alencarliense took command of it and descended to map the course of the river. With him went Ribeiro, the doctor Tanageira, who could no longer walk on account of the ulceration of one foot, three men whom the fever had rendered unable longer to walk, and six men who were as yet well enough to handle the canoe. By the time the remainder of the party came to the next navigable river eleven more fever-stricken men had nearly reached the end of their tether. Here they ran across a poor devil who had for four months been lost in the forest and was dying of slow starvation. He had eaten nothing but Brazil-nuts and the grubs of insects. He could no longer walk, but could sit erect and totter feebly for a few feet. Another canoe was built, and in it Pyrineus started down-stream with the eleven fever patients and the starving wanderer. Colonel Rondon kept up the morale of his men by still carrying out the forms of military discipline. The ragged bugler had his bugle. Lieutenant Pyrineus had lost every particle of his clothing except a hat and a pair of drawers. The half-naked lieutenant drew up his eleven fever patients in line; the bugle sounded; every one came to attention; and the haggard colonel read out the orders of the day. Then the dugout with its load of sick men started down-stream, and Rondon, Lyra, Amarante, and the twelve remaining men resumed their weary march. When a fortnight later they finally struck a camp of rubber-gatherers three of the men were literally and entirely naked. Meanwhile Amilcar had ascended the Jacyparana a month or two previously with provisions to meet them; for at that time the maps incorrectly treated this river as larger, instead of smaller, than the Gy-Parana, which they were in fact descending; and Colonel Rondon had supposed that they were going down the former stream. Amilcar returned after himself suffering much hardship and danger. The different parties finally met at the mouth of the Gy-Parana, where it enters the Madeira. The lost man whom they had found seemed on the road to recovery, and they left him at a ranch, on the Madeira, where he could be cared for; yet after they had left him they heard that he had died.

On the 12th the men were still hard at work hollowing out the hard wood of the big tree, with axe and adze, while watch and ward were kept over them to see that the idlers did not shirk at the expense of the industrious. Kermit and Lyra again hunted; the former shot a curassow, which was welcome, as we were endeavoring in all ways to economize our food supply. We were using the tops of palms also. I spent the day hunting in the woods, for the most part by the river, but saw nothing. In the season of the rains game is away from the river and fish are scarce and turtles absent. Yet it was pleasant to be in the great silent forest. Here and there grew immense trees, and on some of them mighty buttresses sprang from the base. The lianas and vines were of every size and shape. Some were twisted and some were not. Some came down straight and slender from branches a hundred feet above. Others curved like long serpents around the trunks. Others were like knotted cables. In the shadow there was little noise. The wind rarely moved the hot, humid air. There were few flowers or birds. Insects were altogether too abundant, and even when travelling slowly it was impossible always to avoid them—not to speak of our constant companions the bees, mosquitoes, and especially the boroshudas or bloodsucking flies. Now while bursting through a tangle I disturbed a nest of wasps, whose resentment was active; now I heedlessly stepped among the outliers of a small party of the carnivorous foraging ants; now, grasping a branch as I stumbled, I shook down a shower of fire- ants; and among all these my attention was particularly arrested by the bite of one of the giant ants, which stung like a hornet, so that I felt it for three hours. The camarades generally went barefoot or only wore sandals; and their ankles and feet were swollen and inflamed from the bites of the boroshudas and ants, some being actually incapacitated from work. All of us suffered more or less, our faces and hands swelling slightly from the boroshuda bites; and in spite of our clothes we were bitten all over our bodies, chiefly by ants and the small forest ticks. Because of the rain and the heat our clothes were usually wet when we took them off at night, and just as wet when we put them on again in the morning.

All day on the 13th the men worked at the canoe, making good progress. In rolling and shifting the huge, heavy tree-trunk every one had to assist now and then. The work continued until ten in the evening, as the weather was clear. After nightfall some of the men held candles and the others plied axe or adze, standing within or beside the great, half-hollowed logs, while the flicker of the lights showed the tropic forest rising in the darkness round about. The night air was hot and still and heavy with moisture. The men were stripped to the waist. Olive and copper and ebony, their skins glistened as if oiled, and rippled with the ceaseless play of the thews beneath.

On the morning of the 14th the work was resumed in a torrential tropic downpour. The canoe was finished, dragged down to the water, and launched soon after midday, and another hour or so saw us under way. The descent was marked, and the swollen river raced along. Several times we passed great whirlpools, sometimes shifting, sometimes steady. Half a dozen times we ran over rapids, and, although they were not high enough to have been obstacles to loaded Canadian canoes, two of them were serious to us. Our heavily laden, clumsy dugouts were sunk to within three or four inches of the surface of the river, and, although they were buoyed on each side with bundles of burity-palm branch-stems, they shipped a great deal of water in the rapids. The two biggest rapids we only just made, and after each we had hastily to push ashore in order to bail. In one set of big ripples or waves my canoe was nearly swamped. In a wilderness, where what is ahead is absolutely unknown, alike in terms of time, space, and method—for we had no idea where we would come out, how we would get out, or when we would get out—it is of vital consequence not to lose one's outfit, especially the provisions; and yet it is of only less consequence to go as rapidly as possible lest all the provisions be exhausted and the final stages of the expedition be accomplished by men weakened from semi-starvation, and therefore ripe for disaster. On this occasion, of the two hazards, we felt it necessary to risk running the rapids; for our progress had been so very slow that unless we made up the time, it was probable that we would be short of food before we got where we could expect to procure any more except what little the country in the time of the rains and floods, might yield. We ran until after five, so that the work of pitching camp was finished in the dark. We had made nearly sixteen kilometres in a direction slightly east of north. This evening the air was fresh and cool.

The following morning, the 15th of March, we started in good season. For six kilometres we drifted and paddled down the swift river without incident. At times we saw lofty Brazil-nut trees rising above the rest of the forest on the banks; and back from the river these trees grow to enormous proportions, towering like giants. There were great rubber-trees also, their leaves always in sets of threes. Then the ground on either hand rose into boulder-strewn, forest-clad hills and the roar of broken water announced that once more our course was checked by dangerous rapids. Round a bend we came on them; a wide descent of white water, with an island in the middle, at the upper edge. Here grave misfortune befell us, and graver misfortune was narrowly escaped.

Kermit, as usual, was leading in his canoe. It was the smallest and least seaworthy of all. He had in it little except a week's supply of our boxed provisions and a few tools; fortunately none of the food for the camaradas. His dog Trigueiro was with him. Besides himself, the crew consisted of two men: Joao, the helmsman, or pilot, as he is called in Brazil, and Simplicio, the bowsman. Both were negroes and exceptionally good men in every way. Kermit halted his canoe on the left bank, above the rapids, and waited for the colonel's canoe. Then the colonel and Lyra walked down the bank to see what was ahead. Kermit took his canoe across to the island to see whether the descent could be better accomplished on the other side. Having made his investigation, he ordered the men to return to the bank he had left, and the dugout was headed up-stream accordingly. Before they had gone a dozen yards, the paddlers digging their paddles with all their strength into the swift current, one of the shifting whirlpools of which I have spoken came down-stream, whirled them around, and swept them so close to the rapids that no human power could avoid going over them. As they were drifting into them broadside on, Kermit yelled to the steersman to turn her head, so as to take them in the only way that offered any chance whatever of safety. The water came aboard, wave after wave, as they raced down. They reached the bottom with the canoe upright, but so full as barely to float, and the paddlers urged her toward the shore. They had nearly reached the bank when another whirlpool or whirling eddy tore them away and hurried them back to midstream, where the dugout filled and turned over. Joao, seizing the rope, started to swim ashore; the rope was pulled from his hand, but he reached the bank. Poor Simplicio must have been pulled under at once and his life beaten out on the boulders beneath the racing torrent. He never rose again, nor did we ever recover his body. Kermit clutched his rifle, his favorite 405 Winchester with which he had done most of his hunting both in Africa and America, and climbed on the bottom of the upset boat. In a minute he was swept into the second series of rapids, and whirled away from the rolling boat, losing his rifle. The water beat his helmet down over his head and face and drove him beneath the surface; and when he rose at last he was almost drowned, his breath and strength almost spent. He was in swift but quiet water, and swam toward an overhanging branch. His jacket hindered him, but he knew he was too nearly gone to be able to get it off, and, thinking with the curious calm one feels when death is but a moment away, he realized that the utmost his failing strength could do was to reach the branch. He reached, and clutched it, and then almost lacked strength to haul himself out on the land. Good Trigueiro had faithfully swum alongside him through the rapids, and now himself scrambled ashore. It was a very narrow escape. Kermit was a great comfort and help to me on the trip; but the fear of some fatal accident befalling him was always a nightmare to me. He was to be married as soon as the trip was over; and it did not seem to me that I could bear to bring bad tidings to his betrothed and to his mother.

Simplicio was unmarried. Later we sent to his mother all the money that would have been his had he lived. The following morning we put on one side of the post erected to mark our camping-spot the following inscription, in Portuguese:

"IN THESE RAPIDS DIED POOR SIMPLICIO."

On an expedition such as ours death is one of the accidents that may at any time occur, and narrow escapes from death are too common to be felt as they would be felt elsewhere. One mourns sincerely, but mourning cannot interfere with labor. We immediately proceeded with the work of the portage. From the head to the tail of this series of rapids the distance was about six hundred yards. A path was cut along the bank, over which the loads were brought. The empty canoes ran the rapids without mishap, each with two skilled paddlers. One of the canoes almost ran into a swimming tapir at the head of the rapids; it went down the rapids, and then climbed out of the river. Kermit accompanied by Joao, went three or four miles down the river, looking for the body of Simplicio and for the sunk canoe. He found neither. But he found a box of provisions and a paddle, and salvaged both by swimming into midstream after them. He also found that a couple of kilometres below there was another stretch of rapids, and following them on the left-hand bank to the foot he found that they were worse than the ones we had just passed, and impassable for canoes on this left-hand side.

We camped at the foot of the rapids we had just passed. There were many small birds here, but it was extremely difficult to see or shoot them in the lofty tree tops, and to find them in the tangle beneath if they were shot. However, Cherrie got four species new to the collection. One was a tiny hummer, one of the species known as woodstars, with dainty but not brilliant plumage; its kind is never found except in the deep, dark woods, not coming out into the sunshine. Its crop was filled with ants; when shot it was feeding at a cluster of long red flowers. He also got a very handsome trogon and an exquisite little tanager, as brilliant as a cluster of jewels; its throat was lilac, its breast turquoise, its crown and forehead topaz, while above it was glossy purple-black, the lower part of the back ruby-red. This tanager was a female; I can hardly imagine that the male is more brilliantly colored. The fourth bird was a queer hawk of the genus ibycter, black, with a white belly, naked red cheeks and throat and red legs and feet. Its crop was filled with the seeds of fruits and a few insect remains; an extraordinary diet for a hawk.

The morning of the 16th was dark and gloomy. Through sheets of blinding rain we left our camp of misfortune for another camp where misfortune also awaited us. Less than half an hour took our dugouts to the head of the rapids below. As Kermit had already explored the left- hand side, Colonel Rondon and Lyra went down the right-hand side and found a channel which led round the worst part, so that they deemed it possible to let down the canoes by ropes from the bank. The distance to the foot of the rapids was about a kilometre. While the loads were being brought down the left bank, Luiz and Antonio Correa, our two best watermen, started to take a canoe down the right side, and Colonel Rondon walked ahead to see anything he could about the river. He was accompanied by one of our three dogs, Lobo. After walking about a kilometre he heard ahead a kind of howling noise, which he thought was made by spider-monkeys. He walked in the direction of the sound and Lobo ran ahead. In a minute he heard Lobo yell with pain, and then, still yelping, come toward him, while the creature that was howling also approached, evidently in pursuit. In a moment a second yell from Lobo, followed by silence, announced that he was dead; and the sound of the howling when near convinced Rondon that the dog had been killed by an Indian, doubtless with two arrows. Probably the Indian was howling to lure the spider-monkeys toward him. Rondon fired his rifle in the air, to warn off the Indian or Indians, who in all probability had never seen a civilized man, and certainly could not imagine that one was in the neighborhood. He then returned to the foot of the rapids, where the portage was still going on, and, in company with Lyra, Kermit, and Antonio Parecis, the Indian, walked back to where Lobo's body lay. Sure enough he found him, slain by two arrows. One arrow-head was in him, and near by was a strange stick used in the very primitive method of fishing of all these Indians. Antonio recognized its purpose. The Indians, who were apparently two or three in number, had fled. Some beads and trinkets were left on the spot to show that we were not angry and were friendly.

Meanwhile Cherrie stayed at the head and I at the foot of the portage as guards. Luiz and Antonio Correa brought down one canoe safely. The next was the new canoe, which was very large and heavy, being made of wood that would not float. In the rapids the rope broke, and the canoe was lost, Luiz being nearly drowned.

It was a very bad thing to lose the canoe, but it was even worse to lose the rope and pulleys. This meant that it would be physically impossible to hoist big canoes up even small hills or rocky hillocks, such as had been so frequent beside the many rapids we had encountered. It was not wise to spend the four days necessary to build new canoes where we were, in danger of attack from the Indians. Moreover, new rapids might be very near, in which case the new canoes would hamper us. Yet the four remaining canoes would not carry all the loads and all the men, no matter how we cut the loads down; and we intended to cut everything down at once. We had been gone eighteen days. We had used over a third of our food. We had gone only 125 kilometres, and it was probable that we had at least five times, perhaps six or seven times, this distance still to go. We had taken a fortnight to descend rapids amounting in the aggregate to less than seventy yards of fall; a very few yards of fall makes a dangerous rapid when the river is swollen and swift and there are obstructions. We had only one aneroid to determine our altitude, and therefore could make merely a loose approximation to it, but we probably had between two and three times this descent in the aggregate of rapids ahead of us. So far the country had offered little in the way of food except palm-tops. We had lost four canoes and one man. We were in the country of wild Indians, who shot well with their bows. It behooved us to go warily, but also to make all speed possible, if we were to avoid serious trouble.

The best plan seemed to be to march thirteen men down along the bank, while the remaining canoes, lashed two and two, floated down beside them. If after two or three days we found no bad rapids, and there seemed a reasonable chance of going some distance at decent speed, we could then build the new canoes—preferably two small ones, this time, instead of one big one. We left all the baggage we could. We were already down as far as comfort would permit; but we now struck off much of the comfort. Cherrie, Kermit, and I had been sleeping under a very light fly; and there was another small light tent for one person, kept for possible emergencies. The last was given to me for my cot, and all five of the others swung their hammocks under the big fly. This meant that we left two big and heavy tents behind. A box of surveying instruments was also abandoned. Each of us got his personal belongings down to one box or duffel-bag—although there was only a small diminution thus made; because we had so little that the only way to make a serious diminution was to restrict ourselves to the clothes on our backs.

The biting flies and ants were to us a source of discomfort and at times of what could fairly be called torment. But to the camaradas, most of whom went barefoot or only wore sandals—and they never did or would wear shoes—the effect was more serious. They wrapped their legs and feet in pieces of canvas or hide; and the feet of three of them became so swollen that they were crippled and could not walk any distance. The doctor, whose courage and cheerfulness never flagged, took excellent care of them. Thanks to him, there had been among them hitherto but one or two slight cases of fever. He administered to each man daily a half-gram—nearly eight grains—of quinine, and every third or fourth day a double dose.

The following morning Colonel Rondon, Lyra, Kermit, Cherrie, and nine of the camaradas started in single file down the bank, while the doctor and I went in the two double canoes, with six camaradas, three of them the invalids with swollen feet. We halted continually, as we went about three times as fast as the walkers; and we traced the course of the river. After forty minutes' actual going in the boats we came to some rapids; the unloaded canoes ran them without difficulty, while the loads were portaged. In an hour and a half we were again under way, but in ten minutes came to other rapids, where the river ran among islands, and there were several big curls. The clumsy, heavily laden dugouts, lashed in couples, were unwieldy and hard to handle. The rapids came just round a sharp bend, and we got caught in the upper part of the swift water and had to run the first set of rapids in consequence. We in the leading pair of dugouts were within an ace of coming to grief on some big boulders against which we were swept by a cross current at the turn. All of us paddling hard— scraping and bumping—we got through by the skin of our teeth, and managed to make the bank and moor our dugouts. It was a narrow escape from grave disaster. The second pair of lashed dugouts profited by our experience, and made the run—with risk, but with less risk—and moored beside us. Then all the loads were taken out, and the empty canoes were run down through the least dangerous channels among the islands.

This was a long portage, and we camped at the foot of the rapids, having made nearly seven kilometres. Here a little river, a rapid stream of volume equal to the Duvida at the point where we first embarked, joined from the west. Colonel Rondon and Kermit came to it first, and the former named it Rio Kermit. There was in it a waterfall about six or eight feet high, just above the junction. Here we found plenty of fish. Lyra caught two pacu, good-sized, deep-bodied fish. They were delicious eating. Antonio the Parecis said that these fish never came up heavy rapids in which there were falls they had to jump. We could only hope that he was correct, as in that case the rapids we would encounter in the future would rarely be so serious as to necessitate our dragging the heavy dugouts overland. Passing the rapids we had hitherto encountered had meant severe labor and some danger. But the event showed that he was mistaken. The worst rapids were ahead of us.

While our course as a whole had been almost due north, and sometimes east of north, yet where there were rapids the river had generally, although not always, turned westward. This seemed to indicate that to the east of us there was a low northward projection of the central plateau across which we had travelled on mule-back. This is the kind of projection that appears on the maps of this region as a sierra. Probably it sent low spurs to the west, and the farthest points of these spurs now and then caused rapids in our course (for the rapids generally came where there were hills) and for the moment deflected the river westward from its general downhill trend to the north. There was no longer any question that the Duvida was a big river, a river of real importance. It was not a minor affluent of some other affluent. But we were still wholly in the dark as to where it came out. It was still possible, although exceedingly improbable, that it entered the Gy-Parana, as another river of substantially the same size, near its mouth. It was much more likely, but not probable, that it entered the Tapajos. It was probable, although far from certain, that it entered the Madeira low down, near its point of junction with the Amazon. In this event it was likely, although again far from certain, that its mouth would prove to be the Aripuanan. The Aripuanan does not appear on the maps as a river of any size; on a good standard map of South America which I had with me its name does not appear at all, although a dotted indication of a small river or creek at about the right place probably represents it. Nevertheless, from the report of one of his lieutenants who had examined its mouth, and from the stories of the rubber-gatherers, or seringueiros, Colonel Rondon had come to the conclusion that this was the largest affluent of the Madeira, with such a body of water that it must have a big drainage basin. He thought that the Duvida was probably one of its head streams—although every existing map represented the lay of the land to be such as to render impossible the existence of such a river system and drainage basin. The rubber-gatherers reported that they had gone many days' journey up the river, to a point where there was a series of heavy rapids with above them the junction point of two large rivers, one entering from the west. Beyond this they had difficulties because of the hostility of the Indians; and where the junction point was no one could say. On the chance Colonel Rondon had directed one of his subordinate officers, Lieutenant Pyrineus, to try to meet us, with boats and provisions, by ascending the Aripuanan to the point of entry of its first big affluent. This was the course followed when Amilcar had been directed to try to meet the explorers who in 1909 came down the Gy-Parana. At that time the effort was a failure, and the two parties never met; but we might have better luck, and in any event the chance was worth taking.

On the morning following our camping by the mouth of the Rio Kermit, Colonel Rondon took a good deal of pains in getting a big post set up at the entry of the smaller river into the Duvida. Then he summoned me, and all the others, to attend the ceremony of its erection. We found the camaradas drawn up in line, and the colonel preparing to read aloud "the orders of the day." To the post was nailed a board with "Rio Kermit" on it; and the colonel read the orders reciting that by the direction of the Brazilian Government, and inasmuch as the unknown river was evidently a great river, he formally christened it the Rio Roosevelt. This was a complete surprise to me. Both Lauro Miller and Colonel Rondon had spoken to me on the subject, and I had urged, and Kermit had urged, as strongly as possible, that the name be kept as Rio da Duvida. We felt that the "River of Doubt" was an unusually good name; and it is always well to keep a name of this character. But my kind friends insisted otherwise, and it would have been churlish of me to object longer. I was much touched by their action, and by the ceremony itself. At the conclusion of the reading Colonel Rondon led in cheers for the United States and then for me and for Kermit; and the camaradas cheered with a will. I proposed three cheers for Brazil and then for Colonel Rondon, and Lyra, and the doctor, and then for all the camaradas. Then Lyra said that everybody had been cheered except Cherrie; and so we all gave three cheers for Cherrie, and the meeting broke up in high good humor.

Immediately afterward the walkers set off on their march downstream, looking for good canoe trees. In a quarter of an hour we followed with the canoes. As often as we overtook them we halted until they had again gone a good distance ahead. They soon found fresh Indian sign, and actually heard the Indians; but the latter fled in panic. They came on a little Indian fishing village, just abandoned. The three low, oblong huts, of palm leaves, had each an entrance for a man on all fours, but no other opening. They were dark inside, doubtless as a protection against the swarms of biting flies. On a pole in this village an axe, a knife, and some strings of red beads were left, with the hope that the Indians would return, find the gifts, and realize that we were friendly. We saw further Indian sign on both sides of the river.

After about two hours and a half we came on a little river entering from the east. It was broad but shallow, and at the point of entrance rushed down, green and white, over a sharply inclined sheet of rock. It was a lovely sight and we halted to admire it. Then on we went, until, when we had covered about eight kilometres, we came on a stretch of rapids. The canoes ran them with about a third of the loads, the other loads being carried on the men's shoulders. At the foot of the rapids we camped, as there were several good canoe trees near, and we had decided to build two rather small canoes. After dark the stars came out; but in the deep forest the glory of the stars in the night of the sky, the serene radiance of the moon, the splendor of sunrise and sunset, are never seen as they are seen on the vast open plains.

The following day, the 19th, the men began work on the canoes. The ill-fated big canoe had been made of wood so hard that it was difficult to work, and so heavy that the chips sank like lead in the water. But these trees were araputangas, with wood which was easier to work, and which floated. Great buttresses, or flanges, jutted out from their trunks at the base, and they bore big hard nuts or fruits which stood erect at the ends of the branches. The first tree felled proved rotten, and moreover it was chopped so that it smashed a number of lesser trees into the kitchen, overthrowing everything, but not inflicting serious damage. Hardworking, willing, and tough though the camaradas were, they naturally did not have the skill of northern lumberjacks.