The Empirics contended that most compound medicines had been hit upon by chance, and Galen grants that the Dogmatics usually are unable to give reasons for the ingredients of their doses and find difficulty in reproducing a lost prescription.[714] But he holds that reasons can be given for the constituents of the compound and that the logical discovery of such remedies differs from the empirical.[715] His own method was to learn the nature of each disease and the varied properties of simples, and then prepare a compound suited to the disease and to the patient.[716] On the other hand, we see how much depends upon experience from his confession that sometimes he has hastily prepared a compound from a few simples, sometimes from more, sometimes from a great variety. If the compound worked well, he would continue to use it, sometimes making it stronger and sometimes weaker.[717] For as you cannot put together compounds without rational method, so you cannot tell their strength certainly and accurately without experience.[718] He admits that no one can tell the exact quantity of each ingredient to employ without the aid of experience,[719] and says, “The proper proportions in the mixture we shall find conjecturally before experience, scientifically after experience.”[720] In these treatises upon compound medicines, unlike that on medicinal simples, Galen gives the prescriptions of former physicians as well as some tested by his own experience.[721] Sometimes, however, he expresses a preference for the medicines of those writers who were “most experienced”; and once says that he will give some compounds of the more recent writers, who in their turn had selected the best from older writers of long experience and added later discoveries.[722] We suspect, however, that some of these prescriptions had not been tested for centuries. [Sidenote: Suggestions of experimental method.]