Now let us suppose that an apple falls from a considerable height to the ground. Knowing, from former experiences, the hardness of the soil as well as the density of the apple, we can anticipate the effect of the fall. The soil will not show any considerable impression, yet one side of the apple will be crushed. In predicting this result we anticipate sensations that we shall have under a certain set of circumstances. In so far as we shall necessarily have these sensations we have to deal with facts. Not as if our sensations constitute the entire existence of facts; our sensations, being the effects of so-called objective processes upon our senses, are only one end of a relation, which as a matter of course never exists without the other end. Sensations are the one end; they depend upon and vary with the other end. Showing within certain limits as many varieties here as occur there, they represent the other end. We can, and for certain purposes we must, entirely eliminate the subjective and sensory part of our sensations, in order to represent in our minds not how two objects affect our senses of sight or touch but how two or more objects affect each other. Thus we arrive at an objective statement of facts, how the falling apple affects the soil, and the soil the apple; while the relation of both to our senses is to be eliminated. This objective statement of facts is the ideal of all natural sciences. The physicist states the interaction between the falling apple and the soil. He does not care how many sentient beings witness the fall; he does not care about the psychological element in their observations. He abstracts from the subjective elements in their observations as well as in his own, and confines his attention to the objective facts represented in their minds.