The question recurs with remarkable consistence. Not even the wording changes. When I meet casual acquaintances at the office most of them eventually ask, "Are you still canoeing to work?" The question is a cut above the usual elevator chatter, but I do bridle at that word "still," since it implies there is a reason I might stop. For the fourteen years I have been a member of the CIA canoe pool, getting there has been at least half the fun; there is no question of stopping.

Of course, you need special circumstances before canoeing to work becomes feasible, and I am grateful for the way these circumstances have worked out for me. What has made the canoe pool possible is a pair of unwitting conspiracies -- one between nature and the National Park Service, the other between Allen Dulles and a nineteenth-century Marylander named Matthew Ruppert.

The broad, tidal Potomac as it lazes past downtown Washington is familiar to nearly everyone, and fair numbers have seen it as it comes off the Piedmont plateau at Great Falls, a dozen miles upstream. The setting for the canoe pool is the much less familiar stretch between the tidal river and Great Falls. For much of this stretch the salient feature is inaccessibility. The terrain itself is difficult, since most of the way from Great Falls to Georgetown the river runs through a steep-sided gorge, more precipitous on the Virginia than on the Maryland side. But the Park Service (which owns most of the land) has thrown up further barriers in the form of scenic parkways. In Maryland the parkway has turnouts providing access to the Chesapeake and Ohio Canal, which runs parallel to the river. In Virginia, however, the road is a four-lane, almost exit-free speedway that gives motorists no chance to stop and discourages anyone else from crossing it.

To be sure, people walk, bike, and jog by the thousands along the C & O Canal towpath, but almost nobody uses the Virginia shore. Few people, moreover, have any way of using the river itself, and most of those who do are whitewater enthusiasts who stay in the first few miles below Great Falls, where the exciting rapids are. Farther down, the river and the woods on the Virginia side constitute what I have come to think of almost as a private preserve: a narrow strip of near-wilderness, protected by a series of natural and man-made obstacles and used by almost no one.

It was at a point in Virginia on the edge of this stretch that Allen Dulles decided to build the new CIA headquarters, back when nobody worried about the cost of gas and it was the "in" thing to move out of town. My office moved to its new quarters in the fall of 1961, leaving behind "temporary" buildings from World War II that were about to make way for the Kennedy Center. Commuting from Maryland and the District of Columbia was painful at first: the beltway around Washington had not been finished and everyone had to inch across crowded bridges in the District. But three resourceful colleagues of mine discovered the Sycamore Island Canoe Club, set up by Matthew Ruppert and some friends in 1885 and occupying two islands near the Maryland shore almost exactly opposite the Agency building. Thus was both the CIA canoe pool, and its members actually saved commuting time until the beltway was finished.

I did not join the canoe pool until 1970, at first because we lived too far away in the District and then out of inertia after we moved to Maryland. (When we lived in the District my car pool did buy the canoeists a child's edition of Hiawatha and leave it under their canoe.) Since then all three original members have retired and I have become one of the canoe pool's mainstays.

Sycamore Island is an acre or two in area: it lies off the Maryland shore about fifty yards from the canal towpath. On it stands a clubhouse (which has quarters for a caretaker) and an ancient shed full of canoes that often are almost equally ancient. To reach it you descend a short flight of steps from the towpath and yank on a rope that rings a cowbell on the island. This summons the caretaker, who pulls a small, barge-like ferry across to the steps.

From the island over to the Virginia shore, is about a third of a mile. There are no rapids, although the current is brisk when the river rises. The CIA building lies up a steep, wooded bluff and across the parkway, about half a mile from the river.

On a typical morning I wheel my bicycle down our driveway in Bethesda and head for the ferry landing, a little over three miles away. At the landing I meet anyone else who is paddling that day, and we ring for the caretaker and cross to the island. I leave my bicycle in the clubhouse. We unchain our canoe and paddle across. The canoe is left chained to a tree (three canoes have been stolen over the years -- the river is not that remote), the paddles are hidden, and we hike up through the woods on the remnants of a path, complete with stone steps, that was built when the land was a private estate before World War II. We cross the parkway on the cloverleaf that is also carrying our encapsulated colleagues, show our badges to the guards at the gate, and walk through the parking lot to the building. Door-to-door, the operation takes a little under an hour-in the morning; the trip home takes a bit longer because of the uphill bike ride.

"What do you do when it rains?" my elevator acquaintances sometimes ask. The rule is, if it's raining in the morning we drive to work; otherwise we paddle and rely on ponchos if it rains on the way home. But sometimes the rule is hard to apply. Some of us are more eager (or masochistic) than others, and we have been known to have long jesuitical debates by phone over whether given weather qualifies as rain. Afternoon thunderstorms send us scurrying to cadge rides from our office colleagues, and because a poncho is poor protection for a cyclist, I have been known to quail at the sight of a cold winter rain and beseech my friends for a ride.

There are other problems that a non-canoeist might not anticipate. Mud, for example. I have an agreement with my wife that my method of commuting will not add significantly to her laundry pile, and my muddy trousers often test the limits of that agreement. Things are worse in the aftermath of high water, when the emergent riverbank has a consistency almost impossible to describe; Mark Twain's "too thick to drink, too thin to plow" comes close. It is easy to sink in shin-deep. A clothes brush at the office helps, but there still are days when I must try to maintain my dignity with patches of dried mud on the lower third of my trousers.

Or spider-webs. In summer the spiders in the woods on the Virginia shore are incredibly productive. Twenty webs may appear across the path overnight, and some of my colleagues maintain that the spiders lie in wait to throw their webs across our faces. In spider season the first canoeist up the hill carries a branch which he waves in front of him. I am constantly reminded of the "Simba" rebels at the height of the Congo crisis who waved palm fronds to ward off bullets.

These are inconveniences rather than obstacles. Most of the time the annoyances are less severe than the traffic jams on the beltway. I maintain, moreover, that commuting by canoe is far safer than driving; it is just that we have become inured to the dangers of the latter. If conditions are anywhere near normal I think nothing of making the trip by myself. There is not even any particular obstacle to commuting after dark.

A logistic outline like this misses most of the important things, of course. The chance to observe a little corner of wilderness day in, day out is a rare privilege for a deskbound suburbanite. One gets, for example, a glimpse of the long and varied rhythms of the natural world. Deep scars in a tree trunk, gouged by the chain holding my canoe when tree and canoe were submerged in the 1972 flood, are almost healed; where two trees on the hillside have fallen, honeysuckle has used the extra light to move in on the jack-in-the-pulpit; stubby trees on the outcrops in midstream getting stubbier with every flood and ice-storm; fill dirt, dumped when the parkway in Virginia was being built, has been only lightly covered with humus in over twenty years.

Or the seasonal rhythms: the bluebells that come up along the river no matter how much silt has been put down in high water; the unvarying sequence of wildflowers with wondrous names; toothwort, then squirrelcorn, then troutlily; the jopeye-weed festooned with swallowtail butterflies in late summer. I am smug in my knowledge that Dutchman's breeches will appear on a certain patch of ground in April, that a transit of scarlet tanagers can be expected the first week in May, that raspberries will ripen in July and the pawpaws in September. Wood ducks and mallards produce broods every year, orioles invariably nest in the sycamores, and one year we discovered a Carolina wren's nest right beside the path. We wait for the river to change from summer's olive drab to its wintertime gray-blue-green.

There is a long roster of more or less permanent inhabitants: ospreys, herons, pileated woodpeckers, occasional deer, raccoons (rarely seen, but their tracks are everywhere), and the geese -- barnyard and Canada -- that are fed by the caretaker on the island. And beaver: it astonishes my elevator friends that beaver live right in metropolitan Washington, but they are numerous enough to cause a good deal of damage to trees along the bank. They are nocturnal, and it is not unusual to see one swimming home in the morning. On occasion, we have drifted downstream behind a beaver for a hundred yards, and on one evening when I was hurrying through the woods after working late, I nearly tripped over a beaver that was browsing absentmindedly by my canoe.

We have become connoisseurs of sunrises. Spring and fall produce misty scenes out of an oriental painting, and the trees on the Virginia shore explode with color when the autumn sun hits them. On a frigid winter morning we may be spectators to a bleak interplay involving northwest wind, bright sun, dark clouds, and blue sky. Only sodden summer mornings are predictable, and it is during summer that mobilizing the canoe pool is hardest.

It surprises many people, even those familiar with the pressure-cooker effect of a Washington August, that canoeing comes hardest in summer. I sometimes mitigate the discomfort by making the transit in swimming trunks and taking a quick dip. In the morning I still am relatively cool by the time I reach the top of the hill, where I change into business gear just before emerging from the woods. A dip does not help much with the uphill bike ride home, of course, and I must rely for motivation on the prospect of a cold beer.

Winter canoeing is a joy by contrast. Another elevator question is, "Do you mean you canoe all winter?" The answer is emphatically yes -- if the river is not too high, if there is not too much wind, and if the ice is either absent or negotiable. Yes, the canoe pool kept going even in the chilly winter of 1983-84. We obviously must pay closer attention to the many variables involved since the margin for error is drastically reduced, but the variables themselves may combine in ways that offer unexpected opportunities. To take one example, a strong current or a good wind can keep the river ice-free at temperatures well below freezing. To take another, the floating slush that sometimes dots the river slows the canoe but keeps the waves down on a windy day. (Slush on a morning of dropping temperatures is a warning, on the other hand. The original canoe pool once took more than an hour rather than the usual ten minutes to cover the third of a mile back to the island.)

At Washington's latitude, prolonged cold spells are rare and extensive ice is not often the problem. We have had the sustained cold necessary for a solid freeze only three times in the fourteen years I have been paddling. Those three times the winters of 1977, 1981, and 1982, when the river froze all the way across -- we crossed on foot, assuaging the inevitable worries with a wide and creative variety of safety devices: a 14-foot bamboo pole; an ice axe, an automobile towline, so neatly and tightly coiled that it could not conceivably have been used in an emergency; a 15-foot length of rope with the owner at one end and his briefcase bouncing along at the other; and most improbable and useful of all, a canoe. A canoe turns out to be a first class life support system in and on the ice, and with it the pool kept going in conditions that once stopped it.

Walking across when you know the ice is ten inches thick becomes almost comfortable; but what about a thaw? What about the times when the ice builds out from shore, getting thinner and thinner as it approaches open water? I found that a canoe was in its element in these conditions if I brought along an ice axe and a sturdy paddle. Putting part of my weight on the canoe while walking beside it reduced the chance of breaking through and the canoe was a refuge, easy to scramble into, if I misjudged. The ice axe was something useful for chopping a path, but more often for propelling the canoe across the ice in the same way the paddle pushes it through the water. The paddle got the canoe through ice too thin for the axe to grip.

As a result I missed very few days, not just during the depth of the cold spell in 1981, but during the thaw that followed. In 1982, by contrast, I hardly crossed the ice at all. The difference lay in the almost infinite number of variables that affect the quality of river ice. In 1981 the river was exceptionally low, the current was minimal, and the ice formed solidly. In 1982 the ice was more treacherous even though the temperatures were lower. A somewhat higher river level and a heavy fall of snow (which weighed down the ice and let water onto the surface, turning the snow into slush) spelled the difference.

In many ways wind is the most daunting and frustrating of the winter variables. When we are paddling it tries to capsize us, slew us around, and send us sideways all at the same time, and when crossing the ice I have been propelled to the southeast in an almost friction-free environment. A good winter breeze also brings the wind-chill factor down to zero or below, and because in the winter it usually blows straight downstream it reinforces whatever current there is. It was on such a day, when we all were chilled to the bone and there was no way we could stop paddling for even a few seconds, that a colleague remarked he had to keep looking to see if his arms were still there.

For most of the winter, of course, questions of this sort do not arise. Occasionally we get a quiet and beautiful snowfall, which usually occurs when the temperature is around freezing and the river is ice-free. Much more often there simply is nothing to report; there is no ice, the wind is not particularly strong, and the weather is just the normal dreary Washington fare. Dreary or not, I find it more salutary to be out in the weather than to grumble about it from inside; somehow the winter thereby becomes easier to get through. I now do not get really impatient for spring until mid-March.

Aside from the obvious benefits -- the chance to see wildflowers and pileated woodpeckers, the exercise, the insights into the workings of nature - what do I get out of all this? Part of the answer is that regular contact with the earth is as important for me as it was for Antaeus.* Another part (and it may be saying the same thing in a less metaphoric way) is that for a moment I get to evade modern man's almost complete dependence on secondhand information. People now are very largely containerized, physically and even mentally, and without really noticing it we have come to rely on what others tell us about the world beyond our narrow boxes. I suppose this has always been true, but the ratio between the great mass of secondhand data and the small amount we pick up on our own can never have been greater than it is now. It is all to easy to ignore the distinction -- to forget that nearly everything has been through a process of selection, organization, and interpretation before we get to it. This is a particularly serious danger for professional information-processors like me, but I think the proposition holds for most people. At any rate, the canoe commute does give me a firsthand glimpse of what is going on beyond the various manmade containers I inhabit; I benefit from regular access to information that clearly is unmediated.

Beyond that, I find it simultaneously humbling and encouraging to be reminded that it is an endless process out there, always and yet never changing. At a less cosmic level it is satisfying to understand from my own experience (to take just one example) why the Eskimos have a vital need for many terms to distinguish among different kinds of ice. And not least, when things seem to be settling into a pattern of sustained wackiness either at home or in the office, a fixed point of reference like the canoe crossing is useful even if it is brief.

There was only one time when the reference point itself seemed to be working loose. One summer morning, when the mist was still heavy on both the Potomac and my brain, I suddenly noticed that the river was full of dozens of squirming beings ten to twelve inches long, each of them with a huge mouth that stuck above the surface. For a long moment I felt as if I had wandered into a Brueghel painting; then I realized that the surface was covered with insect corpses, the result of some sort of mass death upstream, and the squirming beings were catfish that had come up from the bottom to scoop them in. Relieved that the river was still proof against the absurd, I resumed paddling toward a world I knew was not.

(*In Greek mythology, Antaeus was a giant of Libya, son of Poseidon and Gaea (earth), long invincible in wrestling because his strength was renewed every time he touched the earth, his mother. Hercules held him off the ground and throttled him.)