By the question itself it is, therefore, permitted to form hypotheses, and to select that which shall appear to have the greatest analogy with the other phenomena of nature. But we must exclude from the number all those which supposes the thing already done; for example, such as suppose that all the germs of the same species were contained in the first seed, or that every reproduction is a new creation, and immediate effect of the Almighty's will; because these hypotheses are questions of fact, and on which it is impossible to reason. We must also reject every hypothesis which might have final causes for its object; such as, we might say, that reproduction is made in order for the living to supply the place of the dead, that the earth may be always covered with vegetables, and peopled with animals; that man may find plenty for his subsistence, &c. because these hypotheses, instead of explaining the effects by physical causes, are founded only on arbitrary connections and moral agreements. At the same time we must not rely on these absolute axioms and physical problems, which so many people have improperly made use of, as principles; for example, there is no fecundation made apart from the body, _nulla foecondatio extra corpus_; every living thing is produced from an egg; all generation supposes sexes, &c. We must not take these maxims in an absolute sense, but consider them only as signifying things generally performed in one particular mode rather than in any other. Let us, therefore, search after an hypothesis which has not any of those defects, and by which we cannot fall into any of these inconveniences; if, then, we do not succeed in the explanation of the mechanical power Nature makes use of to effect the reproduction of beings, we shall, at least, arrive at something more probable than what has hitherto been advanced. As we can make moulds, by which we can give to the external parts of bodies whatever figure we please, let us suppose Nature can form the same, by which she not only bestows on bodies the external figure but also the internal. Would not this be one mode by which reproduction may be performed?