After a few general observations on the nature of evidence and the value of human testimony in different cases, he proceeds to define what he means by a miracle. And here at the very beginning of the subject we find that we have to take objection to Hume's definition of a miracle, which exhibits unfounded assumptions and false premises. He gives two definitions in different parts of his essay. The first is, "A miracle is a violation of the laws of nature." The second is, "A miracle is a transgression of a law of nature by a particular volition of the Deity, or by the interposition of some invisible agent." Now both these definitions are bad or imperfect. The first assumes that we know all the laws of nature; that the particular effect could not be produced by some unknown law of nature overcoming the law we do know; it assumes also; that if an invisible intelligent being held an apple suspended in the air, that act would violate the law of gravity. The second is not precise; it should be "some invisible intelligent agent," otherwise the action of galvanism or electricity, when these agents were first discovered, and before they were ascertained to form part of the order of nature, would answer accurately to this definition of a miracle. The words "violation" and "transgression" are both improperly used, and really beg the question by the definition. How does Hume know that any particular miracle is a violation of a law of nature ? He assumes this without a shadow of proof, and on these words, as we shall see, rests his whole argument.

Before proceeding further, it is necessary for us to consider what is the true definition of a miracle, or what is commonly meant by that word. A miracle, as distinguished from a new and unheard-of natural phenomenon, supposes an intelligent superhuman agent either visible or invisible. It is not necessary that what is done should be beyond the power of man to do. The simplest action, if performed independently of human or visible agency, such as a tea-cup lifted in the air at request as by an invisible hand and without assignable cause, would be universally admitted to be a miracle, as much so as the lifting of a house into the air, the instantaneous healing of a wound, or the instantaneous production of an elaborate drawing. It is true that miracles have been generally held to be, either directly or indirectly, due to the action of the Deity; and some persons will not, perhaps, admit that any event not so caused deserves the name of miracle. But this is to advance an unprovable hypothesis, not to give a definition. It is not possible to prove that any supposed miraculous event is either the direct act of God, or indirectly produced by Him to prove the divine mission of some individual; but it may be possible to prove that it is produced by the action of some invisible preterhuman intelligent being. The definition of a miracle, I would propose, is therefore as follows:—"Any act or event necessarily implying the existence and agency of superhuman intelligences," considering the human soul or spirit, if manifested out of the body, as one of these superhuman intelligences. This definition is more complete than that of Hume, and defines more accurately the essence of that which is commonly termed a miracle.