Let me digress here. We have just seen that the geologist possesses an instrument which the mathematician misses and which enables him to conclude from the present to the past. Why does the same instrument not allow us to conclude from the present in the future? If I see a twenty-year-old man, I’m sure he has gone through all the stages from childhood to adulthood, and therefore that there has not been for twenty years on Earth a cataclysm that destroyed all life, but that does not prove to me that there will not be one in twenty years from now. We have, to know, the past of the weapons that we miss when it comes to the future, and that is why the future appears to us more mysterious than the past.
I cannot refrain here from referring to an article I have written on chance; I recalled the opinion of Mr. Lalande who had said, on the contrary, that if the future is determined by the past, but the past is not determined by the future. According to him, a cause can produce only one effect, while the same effect can be produced by several different causes. If it were so, it would be the past that would be mysterious and the future that would be easy to know.
I could not adopt this opinion, but I showed what it could have been the origin. The principle of Carnot shows us that energy, which nothing can destroy, is likely to dissipate. Temperatures tend to equalize, the world tends towards uniformity, that is to say toward death. Large differences in causes therefore produce only small differences in effects. As soon as the differences in the effects become too small to be observable, we have no longer any means of knowing the differences which formerly existed between the causes which gave rise to them, however great these differences were.
But it is precisely because everything tends towards death, that life is an exception that needs to be explained.
If rolling pebbles are abandoned at random on a mountain, they will all eventually fall into the valley; if we find one at the bottom, it will be a banal effect and will not tell us about the previous history of the pebble; we will not be able to know in what point of the mountain it was first placed. But if, by chance, we meet a stone in the neighborhood of the summit, we can affirm that it has always been there, since as soon as it was on the slope, it would have rolled to the bottom; and our certainty will be even greater as the case is more exceptional and it is more likely not to occur.