It will be observed that I can give no positive evidence for the length or breadth of the as. The pieces in the Museum are all fragments, and, even if there were any of them whole, they would not by any means decide the original length, although they would of course represent the weight. For as they are late, they would probably have been made at a time when the original rod was shrinking up into a more compact form, just as the Chinese bronze knives get shorter and thicker. But the fact remains that the as was identified completely with the Roman foot measure, the divisions being the same in each. We therefore may with great probability infer that the as was originally a piece of copper a foot in length, and of a known thickness. We have seen that copper and iron are not weighed in the early stages of society, but are appraised by measurement. Why should not the same hold true for Rome? It may be asked, how came it that the as was taken as the typical unit for weight and superficial measure, and to express even an inheritance? The answer is not far to seek. To express fractional parts has ever been a great difficulty with primitive people. As the Malays cannot conceive abstract numerals, but must append the concrete padi to each of their numbers, so the old Italian found it necessary to employ some concrete object, the subdivisions of which were familiar, to express the fractional parts whether it be of an estate or anything else. The most common unit in use was the rod of copper divided into 12 thumbs. Accordingly, if a Roman wished to say that Balbus was heir to one-twelfth of an estate he expressed this by the homely formula that Balbus had come in for one inch, the denominator 12 being mentally supplied, as everyone knew that there were 12 inches in the copper bar. The same principle of taking some familiar object, the ordinary method of dividing which was known to all men, is seen in the method of expressing one-tenth. The Roman denarius was divided into 10 libellae; accordingly, when Cicero wishes to say that a certain person had come in for a tenth part of an estate he says that he has come in for a libella (heres ex libella). From this the reader will at once see that we might just as well declare that the word denarius is an abstract word meaning unity as make the same assertion about the as. Again, when the Roman land surveyors elaborated their system of mensuration, they found that the simplest method of expressing the fractional parts of the jugerum was to employ the old duodecimal method of the as. Nor is this without a parallel elsewhere. As the yard was the common English unit of linear measure, it was applied to the most common unit of land, the quarter of the hide, which was accordingly termed a yard of land, or a virgate (virga terrae). The English analogy is even still more complete, for as the as or foot-rod became the unit of weight, so in Cambridge the yard of butter is identical with the pound of butter[428].

Our next step will be to trace the process by which the as or rod became the general weight-unit, the pound (libra). The term libra is not the oldest Latin name for weight, for pondus or its cognate verb pendeo, which literally means to hang, is the true claimant for that position. Libra seems properly to mean the balance, as is seen from the legal formula (employed in Mancipatio) per aes et libram, by means of copper and the balance. From the fact that its chief use was to weigh asses of copper, the mass of an as came to be termed the weight par excellence, just as the most usual amount weighed in the Greek talanta (scales) became the talanton par excellence. This process can be illustrated by modern examples. Thus in the south of Ireland potatoes are sold by the unit of 21 lbs., which consequently is termed a weight, and instead of speaking of so many stones or hundredweights, everyone speaks of a weight of potatoes. But, as already remarked, it was only at a comparatively late epoch that the bars of copper were weighed. It would be only with the growth of greater exactitude in commercial dealings that the art of weighing, which was employed for all dealings in gold and silver, would be applied to copper. Just as the Malays and Tibetans have been gradually taught by the careful Chinese to employ weights commercially, so the Italian tribes may have been led to do so under the influence of the astute Greek traders from Magna Graecia and Sicily. The system in vogue for gold was that of our old friend the ox-unit. This is proved from the fact that not only is the oldest gold coinage of the Etruscans, the close neighbours of Latium, based upon this standard, but that also in Sicily and Southern Italy there was the small gold talent, the three-fold of the ox-unit. This three-fold of the stater was also used at Neapolis. Although the earliest Greek colonies in Sicily employed at first the Aeginetic standard for silver, we soon find them reverting to the gold or Euboic standard for that metal, whilst the early silver coinage of the Etruscans (before 350 B.C.) is also of the Euboic standard. We may with high probability assume that when the Sicilians and Italians first essayed to weigh their copper rods, they naturally employed the standard already in use for gold and silver. The highest unit of this was the small talent of 3 staters which weighed about 405 grs. The bar was divided into 12 inches, and it was found that an inch of copper rod closely approximated in weight to the small gold talent. The weight of the bar, which was the ancient unit for copper before weight had been employed, now became the standard weight-unit for that metal. It is to be observed that this ounce of 405 grs., though some 27 grs. less than the full Roman uncia of later times, is only 15 grs. lighter than the Roman ounce prior to 268 B.C., for it is an ascertained fact that the old Roman uncia did not exceed 420 grs.[429] It must be remembered that the weight of the ounce would depend on the standard foot by which the bar was measured. Now, whilst the Roman foot measures 296 millim., there was likewise in use in Campania, and probably in many parts of Southern Italy, a foot of 276 millim. The relation of bars of these lengths and of a given thickness to the Roman libra is not without interest. If we take an ordinary engineer’s table of materials we shall find that a copper rod a Roman foot long, and half a Roman inch in diameter, weighs 5040 grs. Now, as the Roman pound weighs 5184 grs. this approximation seems almost too close to be a mere coincidence. If on the other hand we take a rod of a foot of 276 millim. and with a diameter of the corresponding half-inch, we shall get a pound of 4680 grs. and an ounce of 390 grs, which is certainly not far from the weight of the small gold talent. It follows from this that we may expect pounds of different weights in Italy, according as the foot-unit varies in different districts.

In later times, besides the pound of 12 unciae, there were several commercial pounds on Italian soil, the pound of 16 ounces (from which our own avoirdupois is probably descended), that of 18 unciae, and that of 24. The last two are easy of explanation, since one is simply the double, the other one and a half times the Roman pound. But perhaps a different explanation must be sought for the 16 ounce pound. The foot was divided by Greeks and also by Italians into 16 fingers as well as into 12 thumbs. Was therefore the pound of 16 ounces simply derived from the division of the foot bar into 16 fingers, the weight of the finger being however equated to that of the Roman thumb or inch of copper?

The as, having been once subjected to weight, its hundredfold, the centumpondium or “hundred weight,” became the highest Roman weight-unit. Thus the as and the centumpondium of the Italians correspond to the mina and talent of the Greeks. But it will be observed that the Italians obtained their higher unit by the old decimal system, whereas the Greeks had borrowed the mina and its sixtyfold from Asia. The centumpondium must be regarded as a true-born Italian unit, not one borrowed from Greece or Asia, and of this there is further proof. We saw by the ancient Roman law that the cow was estimated at 100 asses, the sheep at 10 asses. No doubt from time out of mind 100 of the bars of copper, which formed the chief lower unit of barter, made one cow, just as in Annam 280 little hoes make one buffalo (p. 167). When copper came to be weighed, the amount of copper which formed the equivalent of the highest unit of barter, the cow, was taken as the highest weight-unit. From what I have said above it is not improbable that the Roman libra and the Sicilian litra of copper were almost equal in weight. The fact that the Greek writers always employed the Sicilian word litra (λίτρα), to translate the Latin libra, likewise indicates that in the Greek mind there was a tradition of their identity. And if the doctrine here put forward of the original nature of the as be right, nothing can be more likely than that the Italians who had crossed into Sicily and their kinsfolk who had remained behind employed rods of similar size, and that when they began to weigh the latter, the “weight” (libra or litra), derived from the standard copper rod, should be the same in each region, until certain modifications occasioned by new monetary conditions according to the needs of different communities had caused some divergency in coin weights, although as a commercial weight the litra remained unchanged. As Aristotle identified the Aeginetic obol and chalcus with the Sicilian litra and onkia, we may with some plausibility suggest that the ancient Greek copper obol or spike and the Italian as or rod were identical in dimensions and in origin.

In Greece the copper obol rapidly fell in weight, for, when once silver currency had been introduced, copper was thrust aside, and it was not till the fourth century B.C. that copper coins came into use. When the copper obol appears as a coin it is but a small piece, being in fact a mere token.

The history of the degradation of copper was seen better in Sicily, where we found the litra still weighing 990 grs., but it rapidly sank to only 200 grs., evidently in this case also being mere money of account. For as the silver litra was about 13½ grs., unless the 200 grain copper litra was a mere token, silver would have been to copper as 17:1, which is obviously absurd. In the case of the Italian as the process is still clearer, for we have every stage of the as, from the bars which I have described through the libral as (aes grave), the sextantal as, the uncial and half-uncial, down to the small coin of the empire commonly called “a third brass.”

Gold and Silver.

Whilst in the infancy of coining the Sicilian silver litra was probably the same as the Aeginetic obol, that is about 16⅔ grs., the Aeginetic didrachm being probably treated as a decalitron (ten-litra piece), nevertheless after no long time the common Euboic standard of 135 grs. was employed at Syracuse and elsewhere, and we have the authority of Aristotle for the statement that the Corinthian stater was called a decalitron. Corinth, as we saw above, used the 135 grain unit for her famous Pegasi, commonly known as “Colts” (πῶλοι), and therefore the litra was by this time 13½ grs. Now, in Etruria we find about 400-350 B.C. a silver currency struck on this same 135 grs. standard. These coins bear marks of value, 𐌢 on coins of 131 grs., 𐌡 on those of 65 grs., 𐌠𐌠' on those of 32 grs., and 𐌠 on those of 14 and 13 grs. It is plain therefore that the stater of 135 grs. was considered to consist of 10 units of 13½ grs. each. In other words, whatever the Etruscans may have called their stater, it was exactly the same in weight and method of subdivision as the decalitron of Syracuse. At a later period (350-268 B.C.) we find on coins of like weight the symbols 𐌢𐌢 instead of 𐌢, 𐌢 instead of 𐌡, 𐌡 instead of 𐌠𐌠'. The unit now is exactly half of what it was at an earlier stage, 6¾ grs. instead of 13½ grs.

Not till 268 B.C., just on the eve of the First Punic War, did Rome first coin silver. This coin, called denarius, as its name implies, represented 10 asses. It was divided into four parts, each of which was called a sestertius or 2½, and was marked with the symbol 𐆘 representing that number.

It is very remarkable that the Etruscan coin of the second series, marked 2½, is only very slightly heavier than the Roman sesterce (sestertius) which bears a similar mark. Hence it has been very reasonably inferred that when the Romans set about the coinage of silver, they simply adopted with slight modification the silver system employed by their neighbours across the Tiber. This is all the more probable, as it is almost certain that, though Rome did not strike silver she like Athens before the time of Solon, and like Syracuse, used freely the coins of other communities for a long time previously. The Etruscan coins would therefore serve as silver currency at Rome. We may then assume that the monetary system must have been much the same on both sides of the river. Accordingly, since in 268 B.C. we find the Romans striking a coin in silver representing 10 copper asses, which is almost the same in weight as the Etruscan coin marked 𐌢, we may reasonably infer that, if the Romans had commenced coining silver a century earlier, their denarius or 10-as piece would have been the same weight as the Etruscan.

Now besides the litra, which we found to be both a copper-unit and a silver coin in Sicily, there is another term of great interest, especially as it plays an important part in the history of Roman money. The general Latin name for a coin is numus, which in the later days of the Republic usually meant a denarius when used in the more restricted sense, but in the earlier period it was the term specially applied to the silver sesterce (sestertius). This is almost certainly a loan-word, for Pollux is most explicit in warning us that, although the word seems Roman, it is in reality Greek and belongs to the Dorians of Sicily and Italy[430]. It is always a name of a coin of silver in Sicily, being so used by Epicharmus. The coin meant by this poet cannot have been one of great value, for he says: “Buy me a fine heifer calf for ten nomi.” It was in all probability the Aeginetan obol, for Apollodorus in his comments on Sophron set it down at three half (Attic) obols, that is, almost 17 grs. This is confirmed by the fact that an Homeric scholiast makes the small talent weigh 24 nomi, which gives nearly 17 grs. as the weight of that unit. Crossing into Italy, we find that according to Aristotle[431] there was a coin called a noummos at Tarentum, on which was the device of Taras riding on a dolphin. This is the familiar type of the Tarentine didrachms which, from their first issue down to the invasion of Pyrrhus (450-280 B.C.), weigh normally 123-120 grs., although one specimen weighs 128 grs. This coin Mommsen recognized as the noummos of Aristotle. Professor Gardner afterwards suggested that the diobol, on which occasionally the same type is found, was rather the coin meant. Recently Mr A. J. Evans has almost proved this hypothesis impossible by showing that all the diobols yet known are probably later than the time of Aristotle[432]. As, however, this rests on negative evidence, and is liable to be overthrown at any moment by the discovery of an archaic diobol, it is advisable to cast about for some more positive criterion. Heraclea of Lucania, the daughter-city and close neighbor of Tarentum, as we know from the famous Heraclean Tables (which scholars are agreed in regarding as written about the end of the 4th cent. B.C.), employed as a unit of account a silver nomos. It is so probable that the nomos employed at Heraclea (circ. 325 B.C.) would be the same in value as that employed at Tarentum in the time of Aristotle (ob. 322 B.C.), that if we can prove the nomos of Heraclea to be a didrachm and not a diobol, we may henceforth hold with certainty that the nomos of Tarentum was the larger coin.

On the Heraclean Tables it is enacted that those who held certain public land should pay certain fines in case they had failed to plant their holdings properly; four olive trees were to be planted on each schoenus of land, and for each olive tree not so planted a penalty of 10 nomi of silver was to be exacted, and for each schoenus of land not planted with vines the penalty was two minae of silver[433]. The schoenus is identical with the Roman actus (half a jugerum), being the square of 120 feet. Four olive trees were the allowance for each schoenus. Now if we can determine the number of vines which were planted on a schoenus, we shall be able to get a test of the value of a nomos. Two minae of silver contained in round numbers 110 Tarentine didrachms of 123 grs. each, or 675 diobols of about 20 grs. each. Olives were many times more valuable than the vine, so that any result which will make the vine about the same value as the obol will be absurd.

Now Mr A. J. Evans, when in Southern Italy, at my request kindly ascertained that vines, when trained on poles on vineyard slopes, are usually about 3 yards apart, whilst when trained on pollard poplars (as is much more usual in Campagna), they stand about 6 yards apart. In the case of the former about 150 vines would go to a schoenus (1600 sq. yards), whilst in the latter case barely 50. We cannot doubt that the distance between the vines must have been much the same in ancient as in modern times.

If now we take the nomos to be a diobol, each vine is worth 4⅔ nomi, or 14 nomi, according as there are 50 or 150 vines to the schoenus. Now, as the valuable and slow growing olive is only worth 10 nomi, and it is impossible to believe that the relative values of olive and vine could have ever been such as those arrived at on the assumption that the nomos is a diobol, we must turn to the alternative course and take the nomos as a didrachm. The penalty for a schoenus of vines is two minae or 110 didrachms. If 150 vines go to a schoenus, each will be worth about ⅔ didrachm, 15 vines being equal to one olive, or taking 50 vines to the schoenus, each vine will be worth about two didrachms, 5 vines being worth one olive. This result is so rational that we need hesitate no longer to regard the well-known Tarentine didrachm as the nomos (noummos) of Aristotle.

There is such a difference between the nomos of Sicily, identical with the Aeginetan obol, and that of Tarentum that we are forced to conclude that the term nomos is not specially applied to any particular coin unit. In Sicily we found the native unit, the litra, identified in certain cases, at least in earlier times, with the Aeginetan obol as well as with the nomos. Why two names nomos and litra for the same unit? Is one Sicilian and the other Greek? This at least gives a reasonable explanation. The Dorians then in Sicily gave the name to their earliest coins, nomos, with them indicating the unit of currency established by law just as did nomisma among other Greeks. As in Sicily the Aeginetic obol was the legal coin (nomos) par excellence, so at Tarentum, where didrachms were the first coins to be struck, the term (nomos) was applied to that unit. We may therefore expect to find the term nomos applied to various kinds of coins among the Italiotes and Italians, according to the particular coin chosen by each state as its own unit of account.

Accordingly we find the term nomos applied to certain bronze coins struck on the sextantal (two ounce) and uncial standards, at Arpi and other towns, which are inscribed N II (the double nummus), N I (nummus), ..... (quincunx), .... (triens), ... (quadrans), .. (sextans), . S (sescuncia), . (uncia), and Σ (semuncia). The divisions being those of the as, it is clear that the nomos, or current coin in those places, was the reduced as. Finally, when the Romans first use the term nummus, it means the silver sestertius (2½ asses), the one-fourth of the denarius or ten-as piece, which weighed a scruple (i.e. 18½ grs.) at the time of the first Roman coinage of silver. Here we have all our positive evidence for the nomos. As diobols of 18 to 17 grs. are found in the coinages of various towns in Magna Graecia, such as Arpi, Caelia, Canusium, Rubi, and Teate, it has been plausibly held that such a diobol was the nomos par excellence of these states, and that it was from contact with them that the Romans learned both the use and the name of such a monetary unit. But Rome may have been influenced by her Etruscan neighbours, for, as we have seen, the smallest denomination in the second silver series of Etruscan coins (of which the coins weigh 129 grs., 32 grs. and 17 grs. respectively) is just the weight of the Roman sestertius, and bears the symbol 𐌡𐌠𐌠 (2½), just as the latter bears 𐆘 (2½). Taking into consideration these facts, it looks as if the Romans and Etruscans grafted on to a native system the diobol, or current silver coin of Southern Italy, the Romans (and for all we can tell the Etruscans likewise) adopting at the same time the name nummus. Finally, we observe that this nummus is identical with the Sicilian nomos, which in turn was found to be none other than the Aeginetic obol. The Roman sestertius being a scriptulum (17⁷⁄₁₂ grs.) in weight, we thus find a direct connection between the latter and the Aeginetic obol (16⅔ grs.). This need not surprise us, for it is most natural that in the welding of a weight system (partly foreign, and on the native side only employed for gold and silver) and of a system of measurement employed for bronze, certain features derived from the special silver units in use would be introduced into the new system, which afterwards became universal for weighing all commodities. The term Sicilicus[434] employed for the quarter-ounce is good evidence for this hypothesis. Its name seems to mean simply Sicilian. In weight it was about 108 grs. Now, didrachms struck on such a foot are found in the Greek cities of south-western Italy, at Velia, Neapolis and at Tarentum, after the time of Pyrrhus. Did the Romans, who must have carried on by weight all dealings in silver up to 268 B.C., treat such coins as quarter-ounces, and ultimately take the name of the coin (wrongly connecting it with Sicily) to designate the quarter-ounce? In like fashion it was probably discovered that the Aeginetic obol of the Greek colonists was about equal in weight to the line (scriptulum) which is one-twenty-fourth of the inch (uncia) of copper. Thus as there are 24 nomi in the Sicilian talent, so there are 24 scriptula in the Roman uncia. These considerations help to explain the relations which existed between the nomos (Aeginetic obol), sestertius, and scruple.

Mr Soutzo[435] gives a very different account of the nomos. Starting with the Egyptian hypothesis he makes all the Italian weight systems of foreign origin. He thus makes the Roman libra the ⅟₁₀₀ of a Roman talent, which he seems to identify with a light Asiatic talent[436]. Starting with the talent he supposes that on Italian soil it was divided into 100 librae instead of 60 heavy or 120 light minae, as in the East. Each of these librae or pounds was divided into 12 ounces, and each ounce into 24 fractions. He holds likewise that the Italians adopted from the East the use of bronze “comme matière première de leurs échanges,” at the same time as they obtained the first germs of civilization and their first weight standards. The centumpondium or 100 weight therefore he takes as his prime unit. But besides the talent and the mina and the centumpondium and libra or as, according to Mr Soutzo, “all the Italian peoples availed themselves of an intermediate weight unit: this was the nomos or decussis[437]. This unit was the libral nomos, the twelfth of the heavy talent, being worth ten minae or librae, and the libral decussis, the tenth of the centumpondium, weighing 10 librae.” The monetary nomos and decussis, he thinks, played an important part in the history of Italian coinage. He admits however that no specimen of either nomos or decussis of libral standard is known, the heaviest being a decussis of the Roman triental (one-third) standard, whilst the pieces from Venusia and Teanum Apulum marked N I and N II (nomos and double nomos), representing 10 and 20 minas respectively, belong to a still much more reduced standard. The simple multiples of the as (libra) and litra, such as the tripondius and dupondius, were just as rarely cast in the libral epoch. The mina or the as with their fractions, on the contrary, were the kinds most employed: originally the series was ordinarily composed of the as (marked I or sometimes ............), the semis (S), the triens (....), the quadrans (...), the sextans (..), the uncia (.) and semuncia (Σ). In some series the as is rare and the semis is wanting, but in addition to the other denominations here given the quincunx (:·:) and the dextans (S...., 1 semis + 4 unciae) are found. The presence or absence of these pieces characterizes certain Italian and Sicilian monetary systems[438]. All the evidence virtually which can be produced by Soutzo for this hypothetical nomos is that at Syracuse the Corinthian stater of 135 grs. was called a decalitron, that the Tarentine didrachm of 128 grs. (max.) was similarly divided into 10 litras, that the Romans employed the tenfold of the as (decussis) and when they coined silver called their silver unit a denarius as representing 10 copper asses, and the fact that certain copper coins such as those of Arpi, called nomi, were evidently regarded as containing 10 units, the half being the quincunx. But, as we have already seen, the real explanation of these coins seems to be that they represent reduced asses. We must remember that the heaviest Roman as yet known is only 11 ounces, whilst the great proportion of the earliest specimens are only 10 unciae or (dextantals). When the idea of a real copper currency for local purposes gained ground, and it was found that it was not necessary to have the as of account of full weight, and at the same time to enable the state to make a profit of this copper currency which was solely for home use (just as our Mint makes a large profit of our silver coins), the first stage in reduction was to take off an ounce, or much more frequently two full ounces. I have already pointed out the vitality and universality of the uncia as an unit, and have given the reasons for this. Hence arose asses or bars of 10 ounces. The number 10 had of course great advantages, and presently, when further reductions in the copper currency took place, certain communities clave fast to the decimal system and, instead of taking off some more whole ounces, simply reduced the ounce itself, and retained the denomination, continuing to place the marks of value as before. In those Hellenized states of Apulia just referred to this reduced copper as or litra was the legal unit, and therefore denominated a nomos, especially as it probably corresponded in value (at least as money of account) to the silver unit or nomos in circulation in each district. But whilst Mr Soutzo seems wrong in his view of the nomos, there can be no doubt that there was a consensus among the Sicilians and Italians in favour of making an intermediate unit between 1 and 100, the tenfold of the litra and as, into a higher unit. The Syracusan decalitron and the Roman decussis and denarius are incontrovertible facts. For the latter at least a most interesting connection with a unit of barter can be proved. We saw that by the Lex Tarpeia (451 B.C.) a cow was counted at one hundred asses (centussis, centumpondium) whilst a sheep was estimated at 10 asses (decussis). The reader will observe that, even if the theory were true that the Roman centumpondium is the starting-point of the Roman weight system, and that it was borrowed from the East, the cow all the same plays a most important part in the founding of the system. It would be another instance to prove the impossibility of framing a weight standard independent of the unit of barter, just as we have already seen that the Irish, when borrowing a ready-made weight system from Rome, found it absolutely necessary to equate the cow to the ounce of silver, and as Charlemagne had to adjust the solidus by the value of the same animal. If again the centumpondium and as grew up independently as weight units on Italian soil, and copper was weighed there before gold, the cow is evidently the basis of the system; whilst again, on my hypothesis that copper went by bulk in bars of given dimensions, and was not weighed until long after the scales had been employed for gold, the cow is directly connected with that unit of weight (the gold ox-unit of 135 grs.) which ultimately forms the basis of the uncia (as weight) and libra. On every hypothesis alike the cow must be retained as the chief factor in the origin of the Roman weight system. It will be observed that Mr Soutzo offers no explanation why the Romans, instead of retaining the sexagesimal division of the talent which they are supposed to have imported, subdivided it according to the decimal scale. It cannot be alleged that they had any deep-rooted antipathy to the duodecimal system, seeing that the as was divided into 12 unciae, and the ounce into 24 scruples. The fact that the Romans resisted in this respect the Greek influences, which were so potent a factor in their civilization, is strong evidence that the employment of the tenfold and hundredfold of the as was of immemorial native origin, and most intimately connected with the animal units, which must certainly be held to be autochthonous. As we found in Further Asia and Africa hoes or bars of metal as the lowest unit of currency, so many hoes being worth a kettle, so many kettles a buffalo, so in ancient Italy 10 bars (asses) of copper made a sheep, and 10 sheep made a cow. It is exceedingly probable that the same system prevailed among the Sicels and Sicilian Greeks, 10 litras going to the sheep, 10 sheep to the cow. For we saw on an earlier page that at Syracuse down to the time of Dionysius the cow remained the unit of assessment, just as at the present moment the buffalo is the unit of assessment among the villages of Annam; and, just as with the latter the buffalo is the unit of value, so we may well infer that with the Sicilians the cow played the same rôle. It may therefore be assumed with considerable probability that the employment of the decalitron and decussis as monetary units was originally due to their connection with the value of the sheep.

As Soutzo has observed, the degradation of the local copper series moved on most unequal lines, and no doubt in some places the decussis did not represent perhaps one half the value of its archetype, the sheep, whilst at the same moment the copper unit in another community stood at almost its original weight and value. Where silver was coined the degradation of copper went on all the quicker; there was a tendency more and more to get rid of the old cumbrous copper coins, and to employ those of a lighter and more portable size. Moreover the inter-relations between copper and silver made the coinages in these metals act and react upon each other. Thus the state after reducing the copper would reduce likewise the silver, so as to make the two series correspond. This was probably facilitated in some cases at least by the change in the relative value of these metals. Italy was not a silver-producing region, whilst it was rich in copper. Naturally with the increase of commerce and the development of silver mines in neighbouring countries such as Spain, silver became more abundant and the price of copper rose accordingly. We have had occasion already to remark that the abundance or scarcity of gold or silver is indicated by its being employed or not for coinage. In the case of gold we know that it is only when the supply of that metal is in excess of its demand for purposes of ornament that it is or can be employed in the form of coined money. The history of the coinage of Persia, Lydia, Macedonia, Rhodes and elsewhere in ancient times, as well as the history of mediaeval gold coining, make this evident, whilst modern Hindustan teaches us the same lesson. Of course in times of great financial straits under the pressure of war a gold coinage was sometimes issued, as perhaps at Athens[439] in 407 B.C. and as at Rome during the second Punic war in 206 B.C. Backwardness in the coinage of silver among certain peoples is probably to be accounted for in the same way. The employment of iron money at Sparta (and Byzantium) was probably due to the dearth of precious metals rather than to any ordinance of Lycurgus against the employment of the latter. If accordingly we find that Rome did not coin silver until 268 B.C. we are justified in concluding that it was from want of silver she had been so long in following the example of the Etruscans and the Greeks.

It is certainly most significant that within four years after the capture of Tarentum (272 B.C.) and the subjugation of all Southern Italy we find her issuing a well-matured silver currency. Doubtless by her conquests she obtained a vast supply of the precious metal, for we know from the records of Livy and Pliny that great masses of foreign coins and bullion flowed into the treasury after every fresh conquest. We may therefore reasonably assume that previous to 272 B.C. silver had been much dearer in relation to copper.

But to return. We have seen that with the imprinting of some device on the primitive bars of copper, the tendency to reduce their weight would quickly evince itself. Accordingly it was possible that in certain places when the coinage of silver began, and there was still a desire to make the silver unit equal to the copper, the latter having been already reduced, the silver would be proportioned thereto. Thus when silver was first coined in some towns in Sicily, the silver Aeginetic obol of 16½ grs. was regarded as the equivalent of the copper litra, but when Syracuse started a coinage of Corinthian staters, a piece of silver of 13½ grs. was accounted as the litra.

But in other parts of Italy the process was somewhat different. For we find the silver unit when once fixed remaining the same in weight, but simply having its denomination altered to meet the requirements of certain changes in the bronze series. Thus the Etruscan silver staters of the period prior to 350 B.C., which weigh 130 grs., are marked 𐌢, whilst the coins of the same weight at a later epoch are marked 𐌢𐌢, showing that the copper unit had undergone a change. This Soutzo thinks was simply a reduction from the triental to the sextantal foot, and in no wise due to any change in the relative value of silver and copper. That however both influences may have aided in the change will be made clear from the history of the reduction of the Roman denarius and as in the second Punic war. Finally when the Romans coined their first denarii in 268 B.C., the libella or tenth of the denarius, which represented in silver the copper libra, was only 7 grs., an indubitable proof that the as was but then a mere fraction of its former self. Yet all the same it is clear that this silver denarius, which represented a reduced decussis of bronze, had its ultimate source in nothing else than the 10 libral asses which represented the value of a sheep. Are we not then justified in suggesting that the Etruscan stater of 135 grs. marked 𐌢 had a like origin, that the 10 litra piece or noummos of Tarentum of almost the same weight, and the Syracusan 10 litra piece of 135 grs., had also a similar origin, whilst at an earlier period 10 Aeginetic obols (the nomi of the poems of Epicharmus and Sophron) were the equivalent of the same animal? Ten nomi were the price of a calf in the time of Epicharmus, and as we have seen already the value of a sheep and a young calf is always about the same, even down to the present day.

Roman System.

Although it is not our concern to go into the history of Roman money, it is nevertheless necessary to give the reader a short sketch of its principal features in order to make the history of the Roman weight standards intelligible.

First came oxen and sheep, which according to their age and sex bore definite relations to each other, and by which all other values were measured. From an early period (at least 1000 B.C.) copper was in use, not yet however weighed, but estimated by the bulk, as I have already described. Side by side with it ingots of gold and silver passed from hand to hand. Such ingots are mentioned by Varro under the name of bricks (lateres)[440]. Though this mention refers to a later period, we can yet infer from it with certainty that the practice of trafficking in small ingots of gold and silver prevailed in Italy as elsewhere. With gold came the art of weighing, which was also applied to silver. We have given reasons for believing that the weight-unit employed was the same as that which I have termed the ox-unit. We found the Etruscans, the close neighbours of the Romans, and who had access to the gold fields of Upper Italy, employing this unit as their standard from the commencement of their coinage in the 5th century for both gold and silver. Any of the towns of Southern Italy which struck gold, such as Metapontum, coined on the same standard, which was likewise employed for silver, sometimes a little reduced, by many communities, such as Tarentum. The standard ingot of gold would bear a known relation to that of silver, to the bar of bronze, the cow, and the sheep. We have given absolute proof of the relation between cattle and bronze in the 5th cent. B.C., and we may well infer similar constant relations between cattle and bronze, and the other metals. With greater exactness in commercial dealings the bronze rod was next weighed by the standard already in use for gold, and it was found that each of the 12 parts or unciae into which it was divided weighed just three times the ox-unit, that is, the weight of the small talent which we have found likewise in Macedon, Sicily, and Lower Italy, and which may have itself represented originally the conventional value of a slave, which was three cows among the Celts, the close kinsfolk of the Italians, and probably about the same among the early Greeks. As soon as the rods or asses were exchanged by weighing, they would quickly lose their original form, which was only required so long as it was necessary that they should be of certain fixed dimensions. Under the new system it mattered not whether an as was ·8 inches long, and three inches thick, provided only it was of full weight when placed in the scale. These are the pieces which are known as aes rude; as yet they are mere lumps of metal, without any stamp or device. Gaius well describes this stage: “For this reason bronze and the balance are employed (in mancipatio) because formerly they only employed bronze coins, and there were bars (asses), double bars (dupondii), half-bars (semisses) and quarters (quadrantes), nor was there any gold or silver coin in use, as we can learn from a law of the Twelve Tables, and the force and power of these coins depended not on their number but on weight. For as there were bars (asses) of a pound weight, there were also two pound bars (dupondii), whence even still the term dupondius is used, as if two in weight[441]. And the name is still retained in use.” The half-bars likewise and quarters were no doubt proportionately adjusted to weight. It will be observed that the omission of all mention of the decussis as a standard seems to throw additional doubt on Mr Soutzo’s hypothesis. The plain fact is that a mass of bronze ten pounds in weight would have been extremely cumbrous and unhandy for purposes of manufacture into the implements of everyday life.

When and by whom a stamp was first placed on the bars, it is of course impossible to say. Tradition however seems unanimous in assigning it to the Regal period. Pliny’s account of the Roman coinage is as follows[442]: “King Servius first stamped bronze. Timaeus hands down the tradition that aforetime they employed it in a rough state at Rome. It was stamped with the impressions of animals (nota pecudum), whence it was termed pecunia. The highest rating in the reign of that king (Servius) was 120,000 asses, and accordingly this was the first class. Silver was struck A.U.C. 485 (B.C. 268) in the Consulship of Q. Ogulnius and C. Fabius, five years before the first Punic war, and it was enacted that the denarius should pass for ten pounds of bronze, the quinarius for five, and the sestertius for two and a half. Now the libral weight was reduced in the First Punic war, as the state could not stand the expenditure, and it was appointed that asses of the weight of a sextans (2 unciae) should be struck. Thus there was a gain of five-sixths, and the debt was cleared off. The type of that bronze coin was on the one side a double Janus, on the other a ship’s beak, whilst on the triens and quadrans there was a ship. The quadrans was previously termed a teruncius from tres unciae (three ounces). Afterwards under the pressure of the Hannibalic wars in the dictatorship of Q. Fabius Maximus, asses the weight of an ounce were coined, and it was enacted that the denarius should be exchanged for sixteen asses, the quinarius for eight, the sestertius for four; thus the state gained one half. Nevertheless in the soldiers’ pay the denarius was always given for ten asses. The types of the silver were bigae and quadrigae (two-horse and four-horse chariots), hence they were termed bigati and quadrigati[443]. By and by in accordance with the Papirian law half-ounce asses were struck. Livius Drusus when tribune of the Plebs alloyed the silver with an eighth part of bronze. The Victoriatus was struck in accordance with a law of Clodius, for previously this coin brought from Illyria was treated as merchandize. It was stamped with a Victory and hence its name. The gold piece was struck sixty-two years after the silver on such a standard that a scruple was worth twenty sesterces, and this on the scale of the then value of the sesterce made 900 go to the pound. Afterwards it was enacted that 1040 should be coined from gold pounds, and gradually the emperors reduced the weight, most recently Nero reduced it to 45.”

This statement of Pliny is supported in various details by several disjointed passages of Varro and Festus. Thus the former says that “the most ancient bronze which was cast was marked with an animal (pecore notatum)[444], and elsewhere he says that the ancient money has as its device either an ox, or a sheep, or a swine[445],” a statement repeated by Plutarch and other later writers. Festus (s.v. grave aes) says “aes grave was so called from its weight because ten asses, each a pound in weight, made a denarius, which was so named from the very number (i.e. deni). But in the Punic war, the Roman people being burdened with debt, made out of every as which weighed a pound (ex singulis assibus librariis) six asses, which were to have the same value as the former.” We have also a statement in the fragment of Festus (4, p. 347, Müller) that afterwards the asses in the sestertius were increased (i.e. to 4 from 2½), and that with the ancients the denarii were of ten asses, and were worth a decussis, and that the amount of bronze (in the denarius) was reckoned at XVI asses by the Lex Flaminia when the Roman people were put to straits by Hannibal[446]. Again, Festus says: “Asses of the weight of a sextans (two ounces) began to be in use from that time, when on account of the Second Punic war which was waged with Hannibal, the Senate decreed that out of the asses which were then libral (a pound in weight) should be made those of a sextans in weight, by means of which when payments began to be made, both the Roman people would be freed from debt, and private persons, to whom a debt had to be paid by the state, would not suffer much loss[447].” Varro likewise is worth hearing: “In the case of silver the term nummi is used: that is borrowed from the Sicilians. Denarii (were so named) because they were worth ten (coins) of bronze each, quinarii because they were worth five each, sestertius, because a half was added to two (for the ancient sestertius was a dupondius and a semis). The tenth part of a denarius nummus is a libella, because it was worth a libra of bronze in weight, and being made of silver was small. The sembella is half the libella, just as the semis is of the as. Teruncius is from tres unciae; as this is the fourth part of the libella so the quadrans is the fourth part of the as.”