On the whole the former is perhaps the better for our present purpose. As we believe that we have discovered by the inductive method the common unit which lies at the base of all these systems, there is no longer the same necessity for always starting with that which is the less ancient. Besides, if we were nominally to pursue this course, it by no means follows that we would be starting from that which is the best known. _Prima facie_ we ought to start with the Roman system, the tradition of which has remained unbroken down to our own days. We could work back through the system of the Middle Ages to the time of Constantine the Great, from Constantine to the early Empire, and from the Empire to the Republic. Moreover no weight-unit is more accurately known than the Roman pound. But the early history of Rome is so obscure that we have absolutely no records of a time, when Greece had already a literature of a venerable antiquity. Rome has no literary remains and even not more than a very few meagre inscriptions dating from before the first Punic War (263-241 B.C.), the very time when Hellas was already far advanced in the autumn of her life. Then Italy had borrowed so much from Hellas that the enquirer must be cautious as to how far he may be dealing with material of true Italian or merely adventitious origin. As we are concerned rather with the _origin_ than with the later developments of weight-systems, it is plain that for dealing with our principal objects the Italian systems present us with no special aid. The late period (268 B.C.) at which the Romans struck silver coins places us at a still further disadvantage if we start with their system. Greece on the other hand presents us not only with abundant literary records of great antiquity, some of them descending from an age which knew not the uses of coined money, but also with thousands of inscriptions cut in marble or bronze, many of which contain data of great value for dealing with the history of currency and weight, and finally presents us with vast series of coins from which we can learn empirically the coin standards employed in various times and places. But it is the very wealth of material that is in some degree here our difficulty. The special feature of Greek national life was its numerous autonomous states. There was no central authority with a mint which issued coins for a whole empire as was virtually the case in the great Persian kingdom, and at a later period in the Macedonian empire of Alexander the Great. In the palmy days of Hellas each petty state issued its own coinage, following in its silver and copper mintages whatever standard or module it pleased.