Only when all the Smokejumpers in his crew had been accounted for did Dodge fly back to Missoula. It is not hard to visualize him, eyes bloody and clothes dirty, as Sallee found him near the top of the ridge after the fire had passed over him, but it takes a moment of thinking to see him as his wife saw him when he stepped down from the plane in Missoula, fastidious as ever except for the tobacco stains at the corners of his mouth. He had five more years to construct a life out of the ashes of this fire.
Jansson had longer to live than Dodge, but those who knew him say he also had great problems rescuing himself. Asked by the Board of Review at what point he had given up being in charge of the rescue, he replied he just couldn’t remember. He couldn’t remember because he never gave up the charge. For instance, the year of the fire he twice returned to Mann Gulch to check his original observations of the blowup. Afterwards he wrote “Jansson’s Ground Check Statement.” Having twice walked and rerun his route with a stopwatch in hand, he concluded that his present report “is within two minutes of the time I have shown in previous statements.”
In the end, he had to rescue himself from Mann Gulch by asking to be transferred to another ranger district. It had got so that he could not sleep at night, remembering the smell of it, and his dog would no longer come in but cried all night outside, knowing that something had gone wrong with him.
7
PERHAPS JANSSONS GREATEST RESCUE in Mann Gulch occurred later in the year of the fire. Harry Gisborne, the man above all others who made the study of fire a science, was determined to examine Mann Gulch firsthand before winter came and destroyed crucial evidence. His fear of winter was probably accompanied by a fear that he had not long to live, and he had some theories about fire whirls he wanted to test against facts. In particular, he wanted to test a theory he had formed about the cause of the Mann Gulch blowup. So despite a severe heart ailment he was determined to make the trip, and, without letting his close friends or doctor know, he persuaded Jansson, his disciple, to accompany him. Almost literally he was to die for his theory about the cause of the Mann Gulch fire.
The intensity of Gisborne’s interest in the cause of the blowup at Mann Gulch and in blowups in general is still another sign of his being an advance-guard scientist. Even as late as the Mann Gulch fire there was no general agreement about the causes of these explosions of wildfires. A blowup is a phenomenon that occurs rarely and often as unpredictably as it occurred that afternoon in Mann Gulch; to add to its secretiveness, it takes place far from the known habitat of meteorologists and trained weather observers. Throughout history, blowups have been seen almost entirely by survivors of big forest fires, who would not have survived if they had stopped to observe them.
Even though Jansson’s testimony before the Board had described the Mann Gulch fire as “a blowup,” the official Report of Board of Review never uses the noun “blowup” or any such adjective as “explosive.” Discussion of the behavior of the fire is limited primarily to its appearance as a routine fire prior to the crew’s being dropped, perhaps because the Forest Service wanted to downplay the explosive nature of the Mann Gulch fire to protect itself against public charges that its ignorance of fire behavior was responsible for the tragedy. It was not until the 1950s, however, that Clive M. Countryman and Howard E. Graham published articles analyzing fire whirls in wildfires that received general acceptance. And it was only after Laird Robinson and I had taken several trips into Mann Gulch in the late 1970s that we saw clearly how these theories explained the explosive complexity of the Mann Gulch fire.
At the time of the fire or soon after, several of the leading Forest Service scientists stationed in Missoula, such as Jack S. Barrows and Charles E. Hardy, advanced a very different theory as to the cause of the blowup, a theory which still has some standing and from its nature would be very difficult to disprove. This theory is based on the assumption appearing most often when the human mind seeks to explain extraordinary effects—that extraordinary effects must be produced by extraordinary causes. According to this then-prevailing theory, the particular extraordinary cause of the blowup of the Mann Gulch fire was a thunderhead.
Stated simply, this theory presupposes that a thunderhead came along and sat down on the fire—its cool air being heavier than the light, hot air rising from the ground—but the thunderhead never got all the way to the ground in the form of rain. In effect, its sudden weight as it sat down on the top of the fire splattered the fire all around in the form of spot fires, and the gusts of wind that hurry along with dry thunder helped to fan the spot fires and the main fire until in a few minutes it was all fire.
The strongest argument in favor of this theory is that there were highly variable air conditions on this day, which was setting a record for high temperature, and highly variable air conditions and explosions of fire come out of the same bag. The plane ride from Missoula to Mann Gulch had been rough enough to make one Smokejumper get sick and turn in his jumping suit for good. Even Rumsey and Sallee were beginning to feel ill and wanted to be among the first to jump. Furthermore, the pilot reported cumulus cloud formations in the distance at the time the plane was circling the fire, and each of those cumulus puffs signified a heavy updraft of hot air. When columns of hot air reach around twenty-five thousand feet and encounter rain and ice crystals, they are cooled and change to thunderheads. Being now heavier than the hot air around them, they start down but can stop without raining—not, however, without causing a scurry or blast of big winds.
The main trouble with this theory is that none of the survivors mentions a thunderstorm passing by, nor does Jansson, who was in the vortex of the blowup and a casualty of it. Moreover, it is the kind of all-purpose theory you can’t disprove that somebody offers whenever a fire blows.
Of course, there is no way in this cockeyed world of ruling out the extraordinary-cause-for-the-extraordinary-effect. You come by boat to Mann Gulch by way of the cliffs of the Missouri River where extraordinary ocean beds stood up and fought each other, but it seems as if the more that becomes known about big cockeyed things, including the actions of men and women as well as cliffs, the more they seem to reduce to one little cockeyed thing fitting closely to another of the same kind, and so on until it all adds up to one big cockeyed thing. It’s never confusion, though, because ultimately it all fits—it’s just cockeyed and fits and is fire. And of course that is extraordinary.
The extraordinary monster needing explanation is at least in its prenatal form a simple little mechanism. A blowup is a dust kitten that has become a raging monster, but its basic mechanism is that of a swirl of dust that seemingly comes from nowhere and may pick up a loose newspaper and give it a toss. When we think of it as a monster, though, it is natural to think that something out of the sky had to start it spinning, and it is probable that some blowup somewhere was started by a thunderhead making a big wind spinning in circles, and it is proper, in searching for the cause of blowups, to consider the thunderhead theory. But the other basic theory of the origin of blowups, and the one we shall be dealing with, can be called the “obstacle theory.” It is the theory of Countryman and Graham and the theory that has met general acceptance. And, not surprisingly, it is Gisborne’s underlying theory, although he had not developed it sufficiently to explain the Mann Gulch fire correctly. It was not until Laird and I returned to Mann Gulch on hot mornings and continued to puzzle over what we saw there that we began to notice each time the same combination of little things that would fit together to start a fire whirl if, as was the actual case, a fire were present near the mouth of the gulch on its southern side and near the top of the ridge. Among woodsmen there is a preference for causes that are there waiting for you when you return, but admittedly sometimes they drop from the sky.
The obstacle theory in its essential elements is not hard to understand. A wind strikes an obstacle, say a rocky promontory on a ridge, shears off it, and so starts to spin and soon goes into full circles behind the promontory. Any fire caught in these cir
cles will throw off sparks and even burning branches which, if the conditions are right, will start spot fires, and these, when the conditions continue to be favorable, will swell into fire swirls, and when you get caught between them and the main fire you will be as lucky as Jansson if you regain consciousness in time to vomit. All this is easy to visualize if you like to walk by moving waters and note what happens in a stream when it strikes a half-submerged rock or small logjam. The stream shears off it, and the good fishing is where the eddies form on the rear flanks of the obstacle and behind it. No trouble at all for stream fishermen to visualize.
Soon the question of how a strike of lightning in a dead snag high up near the top of a ridge close to the mouth of Mann Gulch became a fire monster consuming Mann Gulch and thirteen elite firefighters turns into the question, Where are the winds of yesteryear? And that poetic question soon turns up the equally poetic answer, Gone with the winds. And that poetic question and answer when translated into direct prose means that you can’t explain the cause of a big fire of long ago if you can’t reconstruct the winds that caused it, and also that nothing is more true than that each individual wind passes and is gone for good. But the practical woodsman, who seldom is a poet, starts with the assumption that at least some of the winds of yesteryear are not gone, if only one knows how to see a wind that is gone. The practical woodsman thinks that he can see a lot of things in the woods that will tell him a lot about what can’t be seen there. For instance, you may already have guessed how much Laird and I explain what we see in the woods by relying on what we have seen when fishing. It shouldn’t be surprising, then, that an important part of our theory of what caused the Mann Gulch blowup was an observation we made from a boat on the Missouri River several miles before getting to the mouth of Mann Gulch.
We were on our first trip together to Mann Gulch, in 1977, and I had been left behind not happily at the mouth of the gulch, where there isn’t much to do or see on a hot August afternoon. Laird clearly and if anything overpolitely had left me behind. He had a theory to check that would take him sidehilling to the head of Mann Gulch, and the unspoken word was that if I went along I would slow him down. He left on a supposedly cheery note to the effect that, while he was killing himself on the hot, bare hillside, I could loaf around the mouth of the gulch, which the river left cool, with plenty of time to find a missing part of the puzzle of what caused the blowup. And, so help me, I more or less did.
Nearly thirty years after a fire has burned over a piece of shale in the Gates of the Mountains, there doesn’t seem to be much to see since almost no trees are left standing; black fallen trees thirty years after don’t seem to offer many opportunities to make contributions to knowledge. I thought to myself, “Maybe you are trying to see something big and important too soon. Maybe it would be surer to come if you tried to work up to it.” So I backed off, with only one slightly odd thing to observe about the mouth of the gulch—that there are a few green and standing trees there, just a short stretch of them, a hundred yards or so of them between the rise and the edge of the old fire, and I thought to myself, “That must have been a hell of a big wind to blow all this fire upgulch after it jumped the canyon. You would have thought a little of the fire would have sneaked a short way backward and toward the river.”
I started walking up the canyon, slowly, very slowly. Above the mouth of the gulch there seemed to be nothing to observe but black fallen trees, and after thirty years of lying on the ground they look pretty much alike. After I reconciled myself to the fact that all I was going to see was black fallen trees, I finally said to myself, “About all that is left for me to see is the way the dead trees fell,” and to my astonishment I just then saw something—or at least something that might be something.
Remember, now, that when I was looking at the way trees had fallen I was really looking for winds that had gone, and almost immediately I saw that the black, dead bodies of the fallen trees on the southern side of the gulch where the fire had started were strangely parallel to each other but at right angles to the top of the ridge. My immediate reaction was Everyman’s reaction. I turned and looked to the top of the ridge on the opposite, or northern, side of the gulch where there were also dead fallen trees—lots of them—and in a pattern, too, but in a puzzling one. They were lying parallel to each other, but, unlike the trees on the southern side, they lay parallel to the top of the ridge. However puzzling the patterns, the patterns had to stand as the remains of winds.
They probably had to stand for prevailing winds and for winds that might still be there. Certainly they had to have been there for some years after the trees burned, long enough for the trees to have rotted in the roots and been blown over. One vast storm might have done it but not likely—the trees couldn’t have rotted uniformly and agreed to topple at the same time. They had to have been worked on fairly regularly over the years. As prevailing winds, they might still be there at their more or less appointed time, although the pattern of winds might since have changed. But it was fairly sure that once and for some years a big wind had blown over the top of the southern ridge and then down it (at right angles to the top of the ridge) and that on the northern side a big wind with some regularity had blown parallel to the ridge near its top.
This was the best I could do until Laird got back from his mission to the head of the gulch, but at first he wasn’t much interested in my report. His own report left him fairly depleted. He told me that we needed a new theory to explain why most of the crew, after leaving the escape fire, kept sidehilling up the gulch instead of going for the ridge. We had imagined a long stretch of impregnable reef blocking their escape. “In fact,” Laird said, “there were several big openings in the reef that they passed by but could easily have crossed through.”
We both felt depleted by this negative report. To spend a day in Mann Gulch, we had had to drag a motorboat on a trailer 130 miles over the Continental Divide just to get to the Missouri River. From there to where we were now in the late afternoon had taken the rest of the day, and it was about time to fold up and start back up the river to get to Missoula not too long after midnight. About all we would be able to show for a long day in Mann Gulch was that one of our theories about the tragedy was proven wrong by the hardest of evidence—the ground. So my report about the mess of burned, fallen trees on opposite sides of the gulch wasn’t going to take away our disappointment over the results of our long day. But it was about all we had to show for it, and I made my report brief. Still, neither of us entirely forgot it. We talked about it at several of our customary lunches in Missoula, and it was not long before the parallel messes began to emerge as something that might be important.
It was only a trip or two later that we started to think of the Missouri River as having a possible connection with the blowup of the Mann Gulch fire. Up to this time, the Missouri River had been scenically interesting to us, but mostly it had meant motor trouble for our boat. We usually spent as much time on the river trying to figure out what the missing parts of our motor were as we did trying to figure out the missing parts of the story of the fire. On this day we had gone nearly a mile before the motor stopped, so we were still roughly five miles from the mouth of Mann Gulch. While Laird was kicking the motor, trying to get it to start a second time, I was trying to size up this piece of the river we were floating on to figure out how I would fish it. But no matter how much I was thinking about something else, even fishing, I was always ready to think about prevailing winds, especially when I got anywhere near Mann Gulch. I had noted a medium-sized wave in the quiet water near shore, and I thought to myself, “That’s funny.”
All I meant by “funny” is that the wave was going the wrong way from the way I thought it should be going or, more exactly, the wind blowing it was going the wrong way for a prevailing wind on a big mountain river at this time of day. This early in the day in hot summer, the prevailing wind on a mountain river should be blowing upstream as follows. The rising sun hits the tops of mountains first. Th
e warming mountain air, being lighter than the cool valley air, will rise, and the valley air will rush upstream to take its place. In the late afternoons or evenings, it is generally the other way around—the tops of mountains cool first, cool air is drawn into the warmer valley below, and the prevailing wind is usually downstream. From here on I kept watching the waves on the river and trying to connect them with Mann Gulch.
I wasn’t able to make that connection immediately, but I was able to find a reasonable explanation for the odd downstream wind that blows in the Gates of the Mountains on hot mornings. The air deep in the Gates of the Mountains is always much cooler than the air on the plains outside the Gates. As the river passes through the cooler air inside the cliffs to the plains, the hot plains air rises and the cool air is drawn downstream to replace it. Accordingly, in the cliffs of the Gates of the Mountains, the prevailing wind in the morning and early afternoon is downstream, just as it is in the late afternoons and evenings when the mountains cool.
Coming downriver and thinking about prevailing winds as I was, you see ahead several great bends in the river so complete they look from a distance as if the river ahead has run into a mountain and disappeared under it. Captain Meriwether Lewis, coming upriver, must several times have wondered if ahead he wasn’t about to run out of river, leaving him and his crew with boats and oars but nowhere to row to. It must have been an overwhelming sight to come suddenly upon what seems to be the abrupt end of a big river several hundred miles before it should end.
The explanation for this mirage only slowly emerges as the mountain ahead under which the river disappears takes shape. Slowly, it emerges as a promontory sticking out into the river and causing the river to make a sharp bend to the northwest. It is an obstacle that turns the river at the mouth of Meriwether Canyon, just upriver from Mann Gulch. The promontory begins at the mouth of the canyon where Lewis and his crew camped overnight and extends half a mile before allowing the river to pass around it. As the river bends north again, the waters swirl with considerable force into the mouth of Mann Gulch just below the promontory. After that, we don’t care anymore where the river goes—to the Mississippi supposedly.