If we were dealing with secular records, I suppose no candid and competent student of history would entertain much doubt that the originals of the three stories and of the two are themselves merely divergent versions of some primitive story which existed before the three Synoptic gospels were compiled out of the body of traditions current about Jesus. This view of the case, however, is incompatible with a belief in the historical accuracy of the first and second gospels.[49] For these agree in making Jesus himself speak of both the "four thousand" and the "five thousand" miracle. "When I brake the five loaves among the five thousand, how many baskets full of broken pieces took ye up? They say unto him, twelve. And when the seven among the four thousand, how many baskets full of broken pieces took ye up? And they say unto him, seven." Thus we are face to face with a dilemma the way of escape from which is not obvious. Either the "four thousand" and the "five thousand" stories are both historically true, and describe two separate events; or the first and second gospels testify to the very words of a conversation between Jesus and his disciples which cannot have been uttered. My choice between these alternatives is determined by no _à priori_ speculations about the possibility or impossibility of such events as the feeding of the four or of the five thousand. But I ask myself the question, What evidence ought to be produced before I could feel justified in saying that I believed such an event to have occurred? That question is very easily answered. Proof must be given (1) of the weight of the loaves and fishes at starting; (2) of the distribution to 4-5,000 persons, without any additional supply, of this quantity and quality of food; (3) of the satisfaction of these people's appetites; (4) of the weight and quality of the fragments gathered up into the baskets. Whatever my present notions of probability and improbability may be, satisfactory testimony under these four heads would lead me to believe that they were erroneous; and I should accept the so-called miracle as a new and unexpected example of the possibilities of nature. But when, instead of such evidence, nothing is produced but two sets of discrepant stories, originating nobody knows how or when, among persons who could believe as firmly in devils which enter pigs, I confess that my feeling is one of astonishment that any one should expect a reasonable man to take such testimony seriously. I am anxious to bring about a clear understanding of the difference between "impossibilities" and "improbabilities," because mistakes on this point lay us open to the attacks of ecclesiastical apologists of the type of the late Cardinal Newman; acute sophists, who think it fitting to employ their intellects, as burglars employ dark lanterns for the discovery of other people's weak places, while they carefully keep the light away from their own position.