The third day was crevasse rescue day. We ascended higher onto the mountain, to perhaps 7000 feet or so, and onto the glacier. This was our first experience on this trip of traveling while roped together.

Climbing onto the more open sections of the mountain gave us a chance to test our navitational ability, using topographic maps, compass, and visible features of the mountain.

It also gave me the first chance on this trip to see something I really enjoy; clouds below us. At one point, we arrived at the top edge of a large canyon. Looking down the valley, we could wath the clouds below us, filling up the valley.

After searching for a while, our guides located an appropriate crevasse; a crack in the glacier that was perhaps 7 ft. (2.5 meters) wide, got narrower as it descended into the glacier (but was deep enough the we could not see the bottom; it was in the blackness), and extended for several hundred feet (at least 100 meters) in each direction.

There are several parameters that affect crevasse rescue: how many people have fallen into the crevasse, whether or not they are able to climb out themselves, how many people are on the surface, and where they are (all on the same side of the crevasse, or some on each side).

The easiest situation is if the person who falls into the crevasse happens to land safely on a ledge and can climb out themselves, and everybody who is on the surface is on the same side of the crevasse. In that case, everyone on the surface can work together to firmly anchor the rope, allowing the person in the crevasse to climb out.

Increasingly difficult situations are if the person in the crevasse can climb out but must rely on the rope, the person in the crevasse is injured and must be pulled out, the party members left on the surface are not all on the same side of the crevasse, or more than one person has fallen into the crevasse.

One by one we climbed down on ropes and hung suspended in mid-air in the creavsse, 10 or 20 feet down. This may sound frightening, and if the rope broke it would be, but it was actually very peaceful hanging there; calm, still, almost as bright as on the surface, not too cold, and very quiet.

In the crevasse, we practiced two situations; climbing out on a rope by using prussiks, and hauling an incapacitated person out.

Prussiks are devices that allow someone to climb out of a crevasse in case they fall into one. They are essentially long loops made of a thin rope. "Prussik" actually refers to the knot used to tie the loop onto the main rope, but the term is commonly used to refer to the loop itself, and to the act of hauling yourself out using such a loop.

In use, the prussik is looped and knotted around the main climbing rope that attaches all of us to each other. The knot is made in such a way that the loop can slide when there no pressure on it, but grabs onto the main rope and will not slide the loop is pulled tight.

There are various techniques for using two such prussik loops to climb a rope. In our case, one loop was attached to our waist, and the other loop had places in which we could put our feet. To climb out of a crevasse using this arrangement, we would stand up in the foot loops (removing pressure from the waist loop), slide the waist loop up on the main rope, put our weight on the waist loop (tightening it up), remove weight from the foot loop (freeing it up), and slide the foot loop up on the main loop. Each such cycle would get us another foot or two up the rope, so that slowly we could climb out.

Well, almost. This works well until we got the the top of the crevasse, where the main rope goes over the lip and out of the crevasse. Invariably, the main rope had dug a foot or two into the snow here, leaving us unable to prussik the last few feet out. Those last few feet were invariably miserable; there's simply no good way to negotiate them. You must then do whatever clawing, scrambling, groping, and hauling you can to get yourself up and over the crevasse lip.

Despite all the effort required to prussik out of a crevasse, it is easy compared to hauling someone out who is unable to get themselves out. Imagine lifting a 200 pound (85 kg) person up on a rope, when the rope has the friction of the snow all along it, plus the extra friction of the edge. The only way to do this is to set up a pulley system that gives at least a 3:1 (and maybe a 6:1) mechanical advantage. We practiced setting up such a pulley system (without letting the stranded person fall further into the crevasse, of course!) and using it.

Whenever we were on a glacier that may have crevasses, we would each travel with our prussiks tied around the main rope just in front of us, and the other end in our pockets. This way, if we should ever fall into a crevasse, we would be ready to climb out without having to fumble (perhaps hanging upside down in the dark), for our prussiks.