To reach the highest mountain in Europe starting, let's say, from Hamburg. I choose Hamburg because it's a pretty low point, it's at the coast and no elevation. There are very minor hills near Hamburg, hills that are berg, which are perhaps 100 metres high also. If you do that in a landscape like the two-dimensional surface landscape of earth, you get stuck there, because in order to continue you would have to make a jump into the Hartz Mountains, to the Brocken. There you would be 1100 metre high or almost 1200 metre high. And I should say that I brought the example of the raindrop running downhill, evolution is a very similar... but we don't say evolution runs downhill, we say evolution runs uphill to the most fittest point. But the mechanics behind it, which you can form in mathematical equations, is a very similar one. So you... this theory all has been carried out and it's been shown. So here we have a mechanism... evolution, selection... provides you with a mechanism to go uphill wherever there's a gradient to go uphill. But if you are near Hamburg at these little hills here, you would have to make a jump of some 50 kilometres or even more to get to the next gradient, that means to the Hartz Mountains. And from the Hartz Mountains you would have to make also jumps of hundreds of kilometres to get to the next higher mountains, let's say to the Great Arber in Czechoslovakia. And from there you have to make another big jump to get to the Alps, and once you are in the Alps your jumpings might become a little bit shorter in order to reach finally the Mont Blanc.

So, in other words on a low dimensional landscape you would have to make very big jumps and the mutations would be too rare. So that's why people found it difficult to get this. Now Peter Schuster could show that in a high dimensional landscape that's very different. You always are close to some sizeable mountain from which you can go on, and if your population is large enough you can do so.