Deduction is the familar form of syllogistic reasoning in which from true premises one can derive necessarily true conclusions by following the rules of deductive logic. If all M are P, and S is M, then S is P. Induction draws conclusions which are not certain from multiple examples. For example, if all observed swans are white, an inductive conclusion is "all swans are white." Or if two-thirds of observed cows are brown, the probability of another cow being brown is assumed to be two-thirds. This is the frequentist definition of probability. If a jar contains 100 balls, 60 black and 40 white, then the probability of drawing a black ball is .6. This is a priori probability and is related to enumerative or exhaustive induction. If you know all the instances of swans in the pond are white, the conclusion "all swans in this pond are white" is true. Abduction as a form of reasoning is relatively new. Charles Sanders Peirce called it abduction to infer a premise from a conclusion. For example, since if it rains, the grass gets wet, one can abduce (hypothesize) that it probably rained. Strictly speaking, abductive reasoning is fallacious, a logical error. But Peirce argued that this kind of reasoning has evolved in humans, who have become adept at selecting the best hypothesis to explain the condition. Peirce identified his abduction with the scientific method of hypothesis-deduction-observation-experiment. In this case, the scientist makes various guesses (hypotheses) to explain some observations. Once the hypothesis is formed, deduction is used to predict other logical consequences. Experiments then establish the truth or falsity of these consequences. Notice that if the deductions predict phenomena not previously known, the confirmed consequences are not a part of the original phenomena that led to the hypothesis (usually inductively). Newly predicted (discovered) phenomena carry more weight than those originally known. Peirce knew that hypotheses need not be arrived at by induction. They could be just intuitions or lucky guesses or, as Einstein later called them, "free creations of the human mind." Their origin does not matter (genetic fallacy). The "truth" of a hypothesis lies in its experimental verification and explanatory power.