We have seen with what freedom Carneades discussed the foundations of morality. It is now evident that in so doing he did not exceed the legitimate functions of criticism. No one at the present day looks on Prof. Bain and Mr. Henry Sidgwick as dangerous teachers because they have made it clear that to pursue the greatest happiness of the greatest number is not always the way to secure a maximum of happiness for oneself. The really dangerous method, as we now see, is to foster illusions in early life which subsequent experience must dispel. With the introduction of practical questions, we pass to the great positive achievement of Carneades, his theory of probable evidence. Intended as an account of the process by which belief is adjusted to safe action rather than of the process by which it is brought into agreement with reality, his logic is a systematisation of the principles by which prudent men are unconsciously guided in common life. Carneades distinguishes three degrees of probability. The lowest is attached to simple perception. This arises when we receive the impression of an object without taking the attendant circumstances into account. The next step is reached when our first impression is confirmed by the similar impressions received from its attendant circumstances; and when each of these, again, bears the test of a similar examination our assurance is complete. The first belief is simply probable; the second is probable and uncontradicted; the third probable, uncontradicted, and methodically established. The example given by Sextus is that of a person who on seeing a coil of rope in a dark passage thinks that it may be a snake, and jumps over it, but on turning round and observing that it remains motionless feels inclined to form a different opinion. Remembering, however, that snakes are sometimes congealed by cold in winter, he touches the coil with his stick, and finally satisfies himself by means of this test that the image present to his mind does not really represent a snake. The circumstances to be examined before arriving at a definite judgment include such considerations as whether our senses are in a healthy condition, whether we are wide awake, whether the air is clear, whether the object is steady, and whether we have taken time enough to be sure that the conditions here specified are fulfilled. Each degree of probability is, again, divisible into several gradations according to the strength of the impressions received and the greater or less consilience of all the circumstances involved.[252]