Shortly before breakfast on July 16, as the men were rowing against a gentle current, one of them spotted near the river bank a cluster of about forty empty "booths," or arbors, that had been built of willow boughs as shelters against the sun. Because the bowers were different from other temporary lodgings the captains had seen along the river, they decided, probably after asking Sacagawea her opinion, that the structures had been put together by some of her people, the Snake or Shoshoni Indians. But where were the builders? The captains had been hoping to meet some of the tribe ever since learning they customarily ranged along the headwaters of both the Missouri and Columbia rivers. 
A meeting had, in fact, been tantalizingly close. Horse sign around the camp suggested the Indians had departed ten or twelve days earlier. At that time, July 4 to 6, the whites had been gathered at the White Bear Islands, separated from the Indian bowers by approximately forty-five river miles—but only sixteen land miles.  White hunters had ranged nearly that far from their base, and Indian food gatherers had probably matched the range. If timing had been right, Shoshoni foragers, armed only with bows, might well have heard the guns of their counterparts. Although the Corps's journalists do not speculate on the matter, it is even possible that the sound of shooting had frightened the Shoshoni into breaking camp.
So very close! Yet there was no use trying to overtake, on foot or in dugout canoes, mounted Indians who had a ten-day start. On the party went, to a buffalo Drouillard had just killed and butchered for their breakfast. There Lewis, experimenting again, ate for the first time "of the small guts of the buffalo cooked over a blazing fire in the Indian stile without preperation of washing or other cleansing and found them very good."
The meal finished, he, Drouillard (who had become Lewis's almost constant companion), and two crewmen started ahead of the dugouts on foot. Their aim was to obtain an astronomical fix on the point, only a few miles ahead, where the Missouri breaks out of the spur of the northern Rockies now known as the Big Belts. It was a pleasant excursion. Lewis liked to walk, and the slopes leading to the gap made by the river were washed with the lemony-gold sheen of blooming cactus and massed sunflowers. He scrambled up a tall pinnacle at the mouth of the canyon and, looking back, marveled at the view of the great herds of buffalo grazing on the plains. Though he did not suspect it, it would be his last sight of those mobile mountains of meat for many, often hungry months to come.
Calculations for latitude made, the four men pushed about five miles into the lower canyon, between walls of hard, black rock some eight hundred feet high. At the foot of a rapid that "roled white over the rocks" (Ordway), Lewis killed an elk for supper. The joy ended after that, for he had forgotten his mosquito netting, and he spent the night flailing at the hordes of insects that attacked him.
The next morning, Clark and the boats overtook the scouts at the foot of the rapid. As a precaution against water damage, the soldiers carried the most vulnerable supplies on their backs to the upper end of the turbulence. The dugouts, double manned, were dragged up with two ropes. The river then became deep, the current slow. Cliffs soared on either side. Tall pines and stunted aspens seemed to rise directly out of the dark, vaulting stone. The constricted bottomlands were choked with sunflowers whose seeds, when mature, could be pounded into flour, currants and serviceberries already "ripe and in great perfection," and gooseberries and chokecherries that soon would be edible. All this Lewis recorded in scientific detail in camp that night, as he did the next morning's sighting of a large herd of bighorn sheep clambering on walls five hundred feet above them, "where we should suppose it impossible for any animal to stand."
About three miles farther on, the fleet came to "a handsome, bold and clear stream, eighty yards wide," which the captains named Dearborn's River, after Jefferson's secretary of war. During the winter, Clark's Hidatsa informants had included the Dearborn's valley with that of the Medicine (Sun) River as furnishing a short way to tributaries of the Columbia. There is no indication in the journals that the captains associated the stream they saw with the one the Indians had told about.  Indeed, the whites were growing increasingly baffled by the disparity between what they were seeing and what they had understood the Hidatsas to say during the winter. The river Indians had not mentioned Maria's River, they had been at least partly wrong about the Great Falls, and they had given no forewarning about the profound gorge the Corps was now threading. And so, as Lewis put it, "we are anxious now to meet with the Sosonees or snake Indians . . . in order to obtain information relative to the geography of the country and also if necessary, some horses."
Horses—if necessary. It is a strange qualification. They had been thinking horses ever since Clark had drawn his conjectural map of the West. But their successful traverse of the Grand Portage had, perhaps, inspired them to believe they could cross the Continental Divide in the same manner, without horses, if the carrying place were no longer than the half-day labor the Hidatsas had predicted. But how trustworthy was Hidatsa information?
The Shoshoni could answer the questions—if the explorers could lay hold of some of that tribe. And so Lewis continued in one of his convoluted sentences, "we thought it better for . . . either Capt. Clark or myself to take a small party & proceed on up the river some distance before [i.e., ahead of] the boats, in order to discover them, should they be on the river, before the daily discharge of our guns, which was necessary in procuring subsistence for the party, should allarm and cause them to retreat to the mountains and conceal themselves, supposing us to be their enemies who reach them usually by way of the river." 
Clark, Joseph Field, John Potts, and York took the first turn at land travel. It was miserable. Uneven masses of small, sharp fragments of rock covered the ridge tops while blankets of prickly pear covered the lower levels. The second night out, sitting beside their campfire, Clark pulled seventeen thorns out of his cut, bruised feet. Because the others had suffered equally, he decided to drop down to the river the next day and rest until the boats appeared. While waiting, he could hunt game for the dugout crews. Locked in the gorge, they might be having trouble getting enough to eat. And after they arrived he could admit, reluctantly, to Lewis that except for a streak of signal smoke he and his scouts had seen (it had been too far off to investigate) the Indian sign he had noticed was too old to mean much. They were still on a blind path.
Back on the river, the boats moved slowly, pulled against the strong current by tow ropes. Then the canyon grew still narrower. Water filled its bottom from wall to wall, leaving no banks to walk on. Willy-nilly the men had to return to their oars and fight a stream thrust almost as powerful as they were. By Lewis's estimate the walls of the gorge rose twelve hundred feet above their heads. Evergreens cloaked the north-facing slope. The cover on the opposite side was thinner and heavily scarred and pitted by erosion. Unable to land in that gloomy place, the rowers labored on until after dark before finding a flat spot big enough to hold their camp. Searching for a suitable name for so singular a place, Lewis called it, with underlining for emphasis, "the gates of the rocky mountains ." 
The next day the mountains receded. By afternoon the river had split into several channels and was winding through a broad plain bound by parallel ranges. Here Lewis noted the double timberlines characteristic of the high valleys of the northern Rockies: a broad belt of evergreens bordered on top by the gray rocks of the "snowey regions" and on the bottom by the treeless, slowly rising floor of the valley. The spacious land looked fine for grazing, but although the men noticed old buffalo bones and excrement, they detected no living buffalo. As if to help make up for the lack of food, Lewis's dog, Seaman, caught several young geese, their wings not yet feathered enough for them to fly, and brought them to his master. (A few days earlier the dog had dragged a wounded deer from the water, killed it, and added it to the larder.)
On July 22, the swift, shallow water of the narrow channels forced the soldiers to resort again to the tow ropes. At times, when the brush was too thick to crash through, they took up their setting poles. These entailed standing in the wobbly dugouts, jamming one end of the pole onto the smooth, slippery stones of the bottom, and forcing the craft upward with all the strength the boatman could muster. Along this stretch, afternoon temperatures, taken always in the shade, rose into the eighties from the pleasant seventies where the mercury had hovered for several weeks.  Even worse was discouragement. At the Gates of the Mountains, to use modern terminology, the river valley had bent not just south but a little east of south, so that their travail was actually taking them away from the Pacific.
Into this misery came sudden rejoicing. As the boats were passing a small creek whose banks showed patches of white earth, Sacagawea made her first gesture of recognition. At this place her people collected material for white body paint. Almost surely, she indicated, the expedition would find Shoshoni a little farther on at a place called the Three Forks. Because the Americans had heard of this key spot during their stay among the Mandans, "This peice of information," Lewis wrote, "has cheered the sperits of the party who . . . console themselves with the anticipation of shortly seeing the head of the Missouri yet unknown to the civilized world."
Clark, too, was buoyed by the news when he heard it from the boatmen at his resting place. Any Indians camped at Three Forks or thereabouts, he thought, would be the ones who had set the smoke signals he had seen. Obsessed with the idea of being the first to meet them, he declined to trade places with Lewis. Since only Joseph Field of his original party was fit to continue by land, he enlisted three fresh men—Joseph's brother, Reuben, Robert Frazer, and Charbonneau, who begged to go and might be able to interpret a little. Then off he limped on July 23, covering twenty-three miles in spite of his fatigue and aching feet. He also killed four deer and left them beside the river for the water travelers. On the 24th he went thirty miles. On the 25th he was even more demanding. That morning he led his men onto a low swell of hills and saw the Three Forks. Left and right and ahead of them a majestic vastness rose to black bands of timber capped by gray, round-topped peaks dappled with snow. Puffs of clouds floated overhead. Below the clouds three streams coiled like dropped pieces of string through sage and grass, willows and rose bushes, but only occasional narrow-leafed cottonwoods. The westernmost stream and the middle one joined first and then drifted on a short distance to pick up the eastern river, where it finished lapping around a gray snout of limestone. He had reached the Three Forks of the Missouri. Beaver sign was everywhere. Otter. Deer. Some antelope. Occasional elk. But no sign of Indians. Whoever had set the smoke signal had vanished. Where?
Clearly the westernmost of the forks reached the high country sooner than the other two did, and so, after impaling a note to Lewis on a stick and thrusting the stick into the marshy ground, Clark took his men twenty-five miles up the gentle valley, traveling west and southwest. When Charbonneau collapsed, they camped, "all of them much fatigued, their feet blistered and wounded by the prickly pear." 
The next morning the interpreter was unable to travel. Leaving one man with him, Clark led the other two on a dry walk through unseasonable heat to the top of a hill twelve miles away. The view was tremendous but revealed no Indians. On the way back the scouts stopped at an icy spring. Clark took what he thought were proper precautions before drinking: he wet his face, hands, and feet in the cold water. The preventative failed him. Almost immediately after gulping copiously, he felt as if he had been slugged. Sore-footed, cramped, and dead tired, he hobbled to the camp where he had left Charbonneau. When Charbonneau said he could walk again, the captain decided to cross the stream they had been following and examine the middle of the Missouri's three forks. They chose as a ford a place where the river was divided by an island. After reaching the island they killed two grizzlies and then waded out into the much deeper water of the second channel. The current was swift and Charbonneau, who could not swim, lost his footing. He might well have drowned if Clark had not dragged him out.
These last exertions almost finished the captain. After forcing himself along for another three miles he caved in, shivering with chills and aching in every muscle. The next morning he staggered on. Arriving at the middle fork, he turned down it, and at three in the afternoon rejoined the main party at the camp Lewis had established near the junction. After he had told Lewis he felt bilious (an old complaint of Clark's) and hadn't had a passage for several days, his friend gave him a dose of Rush's thunderbolt pills and directed some of the men to build a bower where Clark could sleep out of the sun.
Bringing the boats up the river to the forks had been no picnic, either. There were the usual struggles with setting poles and tow ropes. When handling the latter the men often had to leave the brushy banks for the cold, waist-deep water, under a broiling sun. Bearded needle grass added its miseries to the familiar pests of gnats, mosquitoes, and prickly pears. Its barbed seeds, Lewis complained to his diary, "penetrate our mockersons and leather legins and give us great pain until they are removed. My poor dog suffers from them excessively." As for the exhausted men, "I occassionaly encourage them by assisting in the labour of navigating the canoes, and have learned to push a tolerable good pole in their fraize [phrase]."
In spite of all this and amid his worries about Indians, he found time, good Jeffersonian that he was, to jot down a continuing run of natural history observations. He described lovely fields of wild flax that he thought could, in time, be profitably cultivated by pioneering farmers. He was less taken by some wild onions he boiled up for one meal—"strong, tough, and disagreeable." As usual he had a sharp eye for new birds such as ruffed grouse and a colorful woodpecker that later ornithologists would name for him (asyndesmus lewis). On seeing several snakes that sought refuge in the water as the boats neared, he helped the men catch some so that he could examine their teeth to learn whether the fangs were hollow and thus capable of releasing stored poison. (They weren't.) On reaching Three Forks he promptly climbed the limestone bluff around which the eastern branch curled and from that vantage point drew a chart of the meandering streams.
Clark's return on the afternoon of July 27 sealed a decision Lewis had already made. In spite of the dangers inherent in additional delay, the boatmen, weakened "from this continual state of violent exertion," had to have rest and time for making new buckskin moccasins and garments to replace clothing ruined by long periods of immersion in water. Clark, too, badly needed a pause. So the word went around: for the next two days they would stay where they were.
They were not idle days.  While hunters went after deer for venison and skins, Lewis struggled with a problem of definition. Which of the river's three forks was most worthy of keeping the name Missouri? He ruled out the eastern fork, which was the smallest, but the other two were so nearly equal in size that, in his mind, it would be wrong to give either one the preference. This meant thinking up appropriate new names. Consultations with Clark in his bower resulted in their calling the eastern fork Gallatin, in honor of the secretary of the treasury, Albert Gallatin; the middle fork the Madison, after Secretary of State James Madison; and the western the Jefferson, after "the author of our enterprize."
A poignant Indian drama, they soon learned, was associated with the Jefferson. Five years previously, according to Sacagawea, talking through her husband, her family had been encamped with a mobile village of Shoshoni at exactly the spot where the American explorers were then resting. On being warned by lookouts of an advancing war party of Hidatsas, the mountain Indians had retreated up the Jefferson to a sheltering grove of stunted cottonwoods and brush. The Hidatsas discovered the ruse and attacked. In a running fight they killed four men, four women, and a number of boys. As the defeated Shoshoni warriors fled on horseback from the battleground, the rest of the band scampered for safety as best they could. While Sacagawea was splashing frantically across a shoal place in the river, a mounted Hidatsa had swooped her up and had taken her, probably by way of the Yellowstone, to an enforced new home. But, Lewis wrote, she seemed stirred neither by memories of the abduction nor by anticipation of joining her people again. "If she has enough to eat, and a few trinkets to wear I believe she would be content anywhere."
It was a stoicism the captains were having to match in their own way. Certain shocking geographic truths were dawning on them. From what the Hidatsas had said earlier in the year about the 225-mile southwestern course of the Missouri between Great Falls and Three Forks, Clark had estimated the longitude of the latter to be about 117°. To this assumption the captains had attached two known facts: the longitude of the Columbia's mouth was almost 124° and the distance from one meridian of longitude to the next was, in Clark's lexicon, about 41 miles. Seven meridians multiplied by 41 supposedly meant that only 300 or so direct-line miles would separate them, at the forks, from the Pacific.
Matters had not worked out that way, however. Instead of traveling southwest from Great Falls, they had gone a little east of south for about 250 miles. As a consequence they were somewhere between meridians 111° and 112°—or about 500, not 300, direct-line miles from the ocean.  And of course they could not travel in direct lines. Windings like those they had been following from Great Falls would probably swell the distance ahead of them to more than a thousand miles.
There were other considerations. They could no longer be sure they could cross the divide in half a day, even with horses, as the Hidatsas had said. And once the Corps did reach a navigable stream leading to the Columbia, would they find trees suitable for building dugouts or would the land, like the valley of the upper Missouri, be devoid of timber? In addition, there was the problem of altitude. Although the mountains cupping the Gallatin, Madison, and Jefferson rivers did not seem to rise particularly high above the broad valley floor, the appearance was deceptive. Since leaving St. Louis, the expedition had followed the Missouri for close to twenty-five hundred miles and in that distance must have gained almost a mile in elevation. Winters, for anyone caught in the mountains, would be severe, as snow patches still lingering near the summits late in July attested. "Game," Lewis wrote, "may rationally be expected shortly to become scarce and subsistence precarious."
In spite of this bleak prospect, there is no indication the captains thought of turning back or even sitting where they were on the chance a band of Shoshoni would eventually come long with extra horses they would be willing to barter. The explorers had to advance. The Columbia lay ahead and, Lewis wrote, with just a little posturing for whoever might read his journals in the future, "if any Indians can subsist . . . in these mountains with the means they have of acquiring food we can also subsist."
When the Corps resumed its journey on July 30, Lewis again took up his custom of walking off to the sides of the river—the Jefferson now—to see as much of the local flora and fauna as he could while staying more or less abreast of the boats. This day a maze of beaver dams entangled him. He clambered over the obstructions from pond to pond, waded in mud and water to his waist, fought the dense willow brush, and in spite of his labors missed the boats. No worry. He shot a duck for supper, built a fire, made a bed of willow boughs, "and should have had a comfortable nights lodge but for the musquetoes which infested me." 
When the time came to write up his notes, Lewis remarked, as he had at other points along the upper Missouri, on the "vast number of beaver" inhabiting the upper river. He knew fur men would be interested in that and in his recommendation of Three Forks as a site for a trading post (a post would be built there in 1810, but Blackfoot Indians soon put it out of business). What he did not foresee, though the development would not have surprised him, was that at least four of the expedition's men, Peter Wiser, John Potts, John Colter, and George Drouillard, would return to the high, bleak region with parties hunting beaver. Casualties would be high. Indians killed Potts in 1808 and Drouillard in 1810, after horribly mutilating him. Wiser, his tracks befogged by now, found new ways to cross the divide into Idaho, where new bonanzas of fur awaited, only to be killed under circumstances now unknown. Colter escaped a similar fate by outrunning the Blackfeet, who had stripped him stark naked, in a race that became an American legend.  As for little Jean Baptiste Charbonneau, Sacagawea's son—but that's another story, for later on.
After the boats picked up Lewis the following day, his sense of the expedition's predicament grew to urgency. All of the men were on the edge of exhaustion in spite of their rest at Three Forks, and there was neither grog nor, on that day, July 31, fresh meat for reviving their morale and strength. The latter shortage led Lewis to one of his rare criticisms of the crew: "when we have a plenty of fresh meat I find it impossible to make the men take any care of it, or use it with the least frugality. tho' I expect that necessity will shortly teach them this art." 
A portage like the one at Great Falls would not do after all. The Corps must have horses. The next day (August 1, Clark's birthday) Lewis began a new search for the Shoshoni. As companions he chose, in addition to Drouillard, two men who were not in shape to help with the boats. One was Charbonneau, whose ankles still bothered him. The other was Sergeant Gass, who had wrenched his back in a fall. Lewis himself was discomfited by a combined attack of diarrhea and the operation of the salts he had taken to cure it. The day was hot, and they wasted miles on a mistaken shortcut. Still, there were compensations. Lewis saw and described a noticeable new bird, the pinyon jay. He and Drouillard killed two elk and left most of the meat beside the stream as a birthday present for Clark and his hungry boatmen. Curiously, Lewis seems not to have taken his dog, Seaman, along on these exploratory expeditions, perhaps because of the difficulty of feeding the animal but more probably because of the foot-lacerating needle grass and prickly pear cactus that blanketed the ground away from the stream.
During the next two days they struggled with hot days and cold nights, with foot problems on the dry slopes, and, near the stream, with thickets of brush and with marshes caused by overflow from innumerable beaver dams. On August 4, the splitting of the Jefferson into two nearly equal branches confused them briefly. After carefully examining both banks Lewis decided that the one to the left, as he faced upstream, was the proper choice for the boats. He wanted to make sure, however, by first exploring the right (westerly) fork while at the same time acquiring a little more geographic knowledge. He wrote a note to Clark, advising him to take the flotilla up the Jefferson, as he continued calling the left fork, and said he and his party would rejoin the boats somewhere along its course. Sergeant Gass then cut down a willow pole, fastened the note to the green wood, and set up the message beside the stream where he thought the boatmen would be sure to see it. Then off the scouts went, up the right fork toward the mountains, at whose toes they camped.
The next day neither Gass nor Charbonneau was able to continue. Accordingly Lewis sent them back across country to a timbered hill beside the main fork with orders to wait there for him and for the boats. Accompanied by Drouillard, the captain pushed up the sparkling, swift-flowing, but very shallow right fork into the mountains. Finding that it bent due west and then north, the explorers forsook it to scramble up a high point in the foothills. This gave them an aerial view of the wide valley. No question: the Jefferson was the way. It surely held more water than the unnavigable right fork, and it continued southwest—the direction emphasized by the Hidatsas—to a distant gap in the mountains. True, the Hidatsa geographers had not always been right, but where nothing is known with surety any firm statement is a reed to clutch at.
These conclusions reached, Drouillard and Lewis made their way to the camp Gass and Charbonneau had established, arriving two hours after dark. (One has to assume the two invalids had built a large fire as a beacon.) Because they had no food for breakfast the next morning, Lewis sent Drouillard hunting in the triangle between the forks while he went downstream with Gass and Charbonneau to meet the boats. A lucky decision. From a high point Drouillard caught sight of the dugouts toiling up the wrong branch and hurried over to turn them back.
The boats were, in fact, traveling very slowly. The reasons were as they had been for several days—dense brush on the stream-banks, gravel bars that grounded the dugouts, treacherous footing, intense daytime heat and cold nights. The stream curved inordinately, so that each day the fleet traveled many miles but went forward very few. Again and again the soldiers begged Clark to let them travel by land, packing their supplies and essential trade goods on their backs as they had done at the Great Falls. Lacking word from Lewis about what lay ahead, he always refused.
As they neared the forks, they heard Drouillard's voice shouting at them. Soon they saw him, dressed like a native and accompanied not by one but by several Indians curious to see the black marvel and the Shoshoni woman they had heard about. At the same time Sacagawea lost the impassivity that Lewis had thought was one of her characteristics. Dancing with the most extravagant joy, she pointed to the Indians and sucked on her finger to indicate they were of her native band. Delighted by the sights, the Indians began singing boisterously. 
Then one of the riders drew away from the flotilla and galloped back to the camp in Shoshoni Cove, crying that the tab-ha-bones, not the Blackfeet, were on the way. The waiting Indians whooped, Cameahwait hugged Lewis again, and this time Lewis hugged back.