Again, we will have to compromise by having at least some rocker in the keel, the appropriate amount being decided by whether the canoe will see river use, where some maneuverability is essential, and by our experience with paddling a variety of boats, and noting what varying rocker does, and by my personal favorite – stealing numbers from existing boats that are known to perform well. We mustn't forget that a canoe operates at the interface of two fluids. We have discussed hydrodynamic drag, the resistance we experience from the interaction of hull and paddle with water. We also have to cope with aerodynamic drag, which we get whenever there is a wind. If you have paddled into a strong headwind, you quickly realize that the air can be a much more difficult obstacle to overcome than the water. I recall a trip I made on the Ausable River in Michigan, which ends with a paddle across Mio Pond, an impoundment of about two miles length. My companion and I were fighting a fifteen mph wind straight in our faces; it took us three hours to make the two-mile distance. On a calm day it would have taken less than an hour. There is little that can be done to compensate for a direct headwind. It is when the wind is quartering or abeam that the shape of the boat above the waterline can have a major effect. A boat with a high bow and stern (think Indian birch-bark canoe) is very easily blown off-course by winds from the side. Obviously, making your bow/stern as low as possible reduces this problem. Likewise, a boat that is out of trim (riding bow high is the usual problem) will be influenced more by wind. Sea kayaks have bows and sterns which are much lower than the cockpit, which makes them relatively insensitive to crosswinds. In my design, the bow and stern were only about two inches higher than the center; in other words, the line of the gunwale seen from the side was nearly flat. So much for ease of paddling. Let's take a look at stability, which has three aspects – initial stability, terminal stability, and dynamic stability. Imagine you are sitting in your boat on a calm lake. Lean a couple of inches to the left. Does the boat feel like it wants to keep right on leaning, until it dumps you into the water? If so, it has low initial stability. If it hardly reacts at all to your lean, it has high initial stability. Lean a bit more. Are you still dry? See if you can get a little water to come over the gunwale. If you can, and the boat hasn't dumped you, it has high terminal stability. If you are swimming at this point, the terminal stability is low. A hull with very high initial stability feels safe to beginning paddlers; it doesn't seem tippy. Very often though, a stable-seeming boat will quickly shift into low-or- no terminal stability when leaned to its critical point, which may not be all that many degrees away from vertical. A hull that seems quite tippy and unstable when vertical can quickly firm up when leaned fifteen or twenty degrees, and be very resistant to actually going over. The rounder the underwater hull shape, the less stability of any kind (remember the semicircle?), and the boxier (more rectangular) the higher the initial stability (although the shoebox will dump you pretty fast when it hits its critical angle). Another problem with high initial stability occurs when the wind starts kicking up waves, or a powerboat passes nearby. A wave on the beam will cause a strong lean as it passes under the boat, whereas a boat that is initially unstable will hardly react to the wave at all. Now we are moving into the realm of dynamic stability – how does the boat react when it (or the surface it is on) is in motion?