by Daniel B. Botkin, originally published by Perigee Books, a division of Penguin/Putnam, 1999.

This book, originally published as “Passage of Discovery“, is an ecologist’s guide to the first half of the Lewis and Clark trail, their travels up the Missouri River from St. Louis to Three Forks, MT.  I have decided to share this book with the readers of my website, and I am going to present the entire book here, one chapter at a time, with a new chapter appearing each week.  There are more than 40 chapters.  If you follow along and read all of them, you will learn about the entire Missouri River as seen by Lewis and Clark at the beginning of the 19th century, and as I visited it during the 1990s to see what they had seen, and to learn how the countryside had changed.  Comparing what Lewis and Clark saw with what we see today is one of the best insights we can get of how nature and environment in American has changed since European settlement.   I hope you enjoy it and find it rewarding.

- Daniel B. Botkin

All of the chapters published thus far can be found in the Passage of Discovery category. Please note that they are listed in reverse order of date posted.

More books by Daniel Botkin are available for purchase from the Center For the Study Of the Environment bookstore.

37. The Mouth of the Marias River: Which Was the Best Way West?

To the Confluence of the Marias and Missouri Rivers: from Great Falls take Route 87 northeast to Loma (11 miles northeast of Fort Benton).  The highway crosses the Marias River at the confluence with the Missouri.

Near the location upstream that Lewis reached in 1805: from Great Falls take Route 87 to Fort Benton, then take state secondary Route 223 north from that city to where it crosses the Marias River.

On June 3, 1805, the expedition was camped at what we know now as the mouth of the Marias River, but which appeared to Lewis and Clark as the junction of two large rivers.   The problem was that they weren’t sure which of the rivers was the real Missouri — the river that flowed in from the north, or the one that flowed in from the west.  In a certain sense this is arbitrary, because at the confluence of two major waterways, you can call either one the upstream continuation of the main river.  But for Lewis and Clark, the question was: Which river would take them the farthest into the mountains and give them the best route over the Rockies to the Columbia?  The Indians had told them to search for a river that had some great falls on it.  This would lead them to the trails used by the Indians to pass over the mountains.  This is the river they would call the real Missouri.

Choosing the wrong river and following it would have serious consequences for the expedition.  “To mistake the stream at this period of the season, two months of the traveling season having now elapsed, and to ascend such stream to the Rocky Mountains or perhaps much further before we could inform ourselves whether it did approach the Columbia or not, and then be obliged to return and take the other stream would not only lose us the whole of this season but would probably so dishearten the party that it might defeat the expedition altogether,” Lewis wrote.

Although the future of the expedition depended on the right choice, even perhaps their lives,  Lewis took a detached, almost scientific approach, as if he were a modern scientist sitting in a comfortable laboratory office, rather than at a rough camp in bad weather. He pursued the problem with a seeming academic curiosity, writing.  “An interesting question was now to be determined: which of these rivers was the Missouri.”

In a sense, the expedition was lost and needed directions.  We are all familiar with this problem.  Perhaps you are lost at this very moment that you read this section of my book, trying to follow the directions at the top of this section.  But the trouble was, in Lewis’s time, there wasn’t anybody handy to give them directions.  Today you can stop at a gas station, use your car phone if you have one, or even rely on a GPS device that is in some automobiles, to give you directions.

The problem the expedition faced was one of their own uncertainty.   The rivers were set in their directions and were not about to move at random over the next few days.

And so Lewis proposed an experiment: “To this end an investigation of both streams was the first thing to be done.”  He recognized the need to measure things about the river, making observations quantitative, to “learn their widths, depths, comparative rapidity . . . and thence the comparative bodies of water furnished by each,” and by these means attempt to infer which was the main stream.

Like a modern scientific team, the camp divided into two groups, each examining the available evidence and each proposing what we would refer to today as an hypothesis.  Most of the men believed the north fork was the main river and therefore the one to follow.  Lewis reviewed the evidence on their side:  The north fork was deeper but not as swift.  However, its waters ran “in the same boiling and rolling manner which has uniformly characterized the Missouri throughout its whole course.”  The waters were brown, thick, and turbid – the big muddy, so it seemed.  The bed of the river was also mainly mud, so that the “air and character of this river” seemed “precisely that of the Missouri below.”  For these reasons, most of those on the expedition were convinced that the north fork was the Missouri.  On the other side were Lewis and Clark who, Lewis wrote, were “not quite so precipitate.”

They decided to explore both forks, what scientists would call testing the two hypotheses.  The next morning Clark led a group up the left fork, while Lewis took a group on the right.  The rest of the expedition remained at the base camp where the two forks joined.  Lewis traveled up the north fork from June 4th to the 6th.  He found that this fork continued northward toward what is now the border between Montana and Alberta, Canada, and he became convinced that this direction was too far to the north to be the route to the Pacific.

After taking time to attempt a reading of the latitude and longitude (which failed because of cloudy weather) he began his return on June 7th to the junction of the two forks to rejoin the main body of the expedition.  Lewis was correct; the north fork was a small tributary that they named the Marias River (actually Maria’s River, in honor of Lewis’s cousin, Miss Maria Wood, but after a while, people dropped the apostrophe).

By waiting a few extra days on the Marias River to try to take measurements to determine his latitude and longitude, he was trying to reduce the uncertainty about the position of the expedition.  But a change in the weather, something he could not make accurate predictions about, prevented him from making the measurements.

In deciding which was the right river, Lewis and Clark were confronted simply with a lack of information.  They were uncertain about what to do and wanted to avoid making a crucial error. The error they faced at the junction of the two rivers was what scientists call an error of uncertainty of the first kind – a problem about the facts of a situation that already exists, or, given present conditions, must occur.  One of the channels was the main river – a fact that was not going to change during the time of the expedition.  There was only one correct river to take.   There was something direct and simple to do to resolve this uncertainty – explore the two rivers and determine by direct observation which was the correct one.

There is another kind of quality about nature that leads to a lack of certainty about what we can do.  Lewis experienced this kind on his way back to the confluence of the Marias and Missouri River.  This is the problem that we know that certain kinds of events can happen, but we don’t know when.  This is an uncertainty of the second kind – uncertainty of the occurrence of some event that has some probability of happening, but whose occurrence involves inherent uncertainty.

I call this is the Las Vegas uncertainty: Will you place a bet on dice that haven’t been rolled yet?  Unlike the first kind of uncertainty, the second kind is not resolved so directly and simply.  You can’t pick up your car phone, call the weather bureau and ask to know with complete certainty whether a thunderstorm will strike Loma exactly where the two rivers come together.  The best a weatherman can do is give you the odds on whether or not it will happen.  We cannot reduce the uncertainty of this kind of future event by studying it.  This is the problem with the flooding on the Missouri River.  It is an uncertainty of the second kind that leads us to build levees and dams.

We can, however, learn what the odds – or at least get an estimate of the odds – and decide if we want to accept those odds. We know, because people have rolled dice for a long time, and also because of mathematical analyzes of probabilities, what the chances are of any number coming up with a legitimate pair of dice, and we know that the number seven is most likely.  But we can’t find a route to always getting the number seven, the same way we can take Route 87 to Loma.

Traveling on Route 87, the countryside appeared similar to the badlands that Lewis and Clark described along the Missouri breaks, near to where we were driving.  Another historic marker told us that we had reached Marias River, “The Lewis and Clark expedition camped at the mouth of this river just east of here, June 3, 1805.”  A land of good directions, maps, and apparent certainty.

But Lewis and Clark saw it very differently.  Pushing up the river in a slow-moving boat against a six-mile-an-hour current, sleeping on its banks, studying the land through which it flowed, Lewis and Clark saw a Las Vegas style river.  Perhaps our problem with this kind of error in our knowledge is a matter of relative time scales.  Lewis and Clark spent more than a year on the Missouri, a time as long or longer than some of the variations of the river.  In this day of satellite and aircraft observations, of automobile travel, and vacations that are quick stops here and there, our time of observation is much shorter.  Most of us have just one shot to see the Missouri River of Lewis and Clark.  If it is flooded, well, we may lose that chance.  If it’s a dry year, we will remember the river as it looked that year and it will be fixed in our imagination as if it were always and forever that way.

On Lewis’s return, following the Marias River downstream, it began to rain, and the peculiar clay soil of the floodplain turned into a slippery mess, difficult to traverse.  After a “most disagreeable and restless night” camped in the rain, Lewis and his small band set off down river to join the rest of the expedition.  The clay soil prevented the rain from soaking through and became so slippery that it was like “walking over frozen ground which is thawed to small depth.”  We know today that they were walking on a clay derived from glacial till and shale, commonly called gumbo, a clay that turns into a plastic and sticky material when wet.

Lewis slipped on this soil while walking on a bluff above the river, but managed to save himself from falling 90 feet to the water.  Just after he had saved himself, he heard one of his men, Windsor, cry out “God, Captain, what shall I do?”  Lewis saw that Windsor had slipped on the clay and slid so that his right arm and leg hung over the bluff and he was holding on to the edge with his left arm and leg.  “I expected every instant to see him lose his strength and slip off,” Lewis wrote, but “I disguised my feelings and spoke very calmly to him and assured him that he was in no kind of danger.”  Lewis then astutely told Windsor to take his knife out of his belt with the hand that was hanging over the precipice, and dig a hole in the bank for his right foot, and by such effort work his way up, which Windsor did, and in that way he was saved.

Searching for the right fork is an inherently different problem from trying to avoid slipping on wet clay and falling into a river.  The second kind of uncertainties are referred to today as problems of risk, because the event has not yet happened and its occurrence has to do with inherent chance, or with processes whose causes, for all practical purposes, we cannot distinguish from true chance events.  Translated into human events, risk becomes a matter of prediction, forecasting, luck, and fortune, the latter two of which were also constant companions of the expedition.

Our modern environmental problems confront us with both kinds of uncertainty.   And it’s important that we understand which kind we are facing.  The floods on the Missouri in the 1990s showed us that we cannot treat uncertainties of the Las Vegas kind as if they were uncertainties of route directions.  We do not seem to have trouble accepting the idea of our own errors – that we might not know which river to take.  But we have a great difficulty understanding and accepting the second kind of uncertainty – that there may be some inherent chance in nature.

On the level plain we passed fields of hay and wheat, and we saw mountains along the horizon.  About 46 miles from Great Falls, we passed over a bridge where we saw the Marias River.   Here the river appears as a small meandering current in a floodplain of cottonwoods.  I think about the irony of what geographers say today: If Lewis and Clark had followed Marias River, they would have found a better pass through the mountains, one eventually used by the railroad and crosses the divide in Wyoming, that would have taken them more easily to the Columbia River.

If you have the time to take a canoe trip through the wild and scenic portion of the Missouri, you may have a chance to experience the river at the Lewis and Clark time scale.  When a friend of mine did this during a five-day trip, he got caught in an intense thunderstorm, and experienced the second kind of uncertainty directly.  Another friend canoed the region slightly upriver from Loma, and he was caught in a strong easterly wind, a headwind, so he and his companion had to canoe hard against that wind in spite of the fact that the river was flowing with them.

The Missouri’s refusal to stay put and stay constant has been the source of many a good story and pithy saying, but this quality has also interfered with our society, with commerce, and with our conservation of nature.  Most of our past methods to conserve and manage environmental factors assume the constancy of nature – except for human intervention.  But the reality is the other way around.  We try to fix a natural varying environment, believing that our interventions are the causes of variations in an otherwise static structure of environment. Like the fickle Missouri, all of nature changes at many scales of time and space.  We have longed for and tried to create an environment that is fixed, like the channelized Missouri downstream.  Having lost our heritage about the river and the prairie, we seem to have ignored its important message.