It is always dangerous to deal with things Keltic. So much difficulty is there in getting at any facts amidst masses of wild assertions and loose conclusions, that a prudent man may well shrink back. However, as it is worth while to give some facts respecting the actual weights of gold rings and other ornaments, I have thought it best to print the following pages.

Attempts have long ago been made to find the standard of the so-called ring money. Sir William Betham, followed by John Lindsay[450], after weighing many examples, arrived at the conclusion that they are based on the ounce Troy. Now as the ounce Troy is entirely unknown to the Brehon Laws, and was only brought into Ireland by the English settlers, it is needless to argue further against that doctrine. Dr Petrie’s[451] discussions about Irish coins are similarly vitiated by his treating as Troy grains the grains of wheat mentioned by the authorities.

1. Irish. Let us work back from the known to the unknown.

The system in the Brehon Laws is as follows:

Unga = 576 grs. of wheat.

The ounce seems to be the highest unit of weight, and just as in the Brehon Laws an unga of silver is equated to a cow, so in early times an unga of gold seems to have been the regular value of a slave, the most valuable of living chattels. At least we may so infer from a curious story of St Finnian of Clonard:

I am indebted for this valuable reference, which also enables us to form an idea of the relative value of gold and silver in early Ireland, to the Rev. B. Mac Carthy, D.D., of Youghal.

But there is another weight called crosoch (crosóg or crosach), found in the most ancient poems. For instance in Cuchulaind the brooch of Queen Medbh, “My spear brooch of gold which weighs thirty ungas, and thirty half ungas, and thirty crossachs and thirty quarter [crossachs].” (O’Curry, Manners and Customs, Vol. III. p. 102.) The weight of a crosoch we learn from a gloss quoted by O’Donovan (Supplement to O’Reilly’s Dictionary) from MS. R. I. A., No. 35, 5. 49.

“Two pinginns and a fourth of a pinginn are a milk of a sheep, i.e. a crosóg.” Since 1 pinginn = 8 grs. wheat therefore a crosóg = 18 grs. wheat or 13·5 grs. Troy.

There are accordingly 32 crosochs in the unga of the Brehon Laws.

Inspection at once shows that the crosoch must have belonged to a different system, on which either the system of ungas and screapalls was grafted or vice versa. The expulsion of the crosoch from the later Irish shows that the first alternative is the true one.

Again, it is certain that the unga and screapall were borrowed from the Roman system, probably before the time of Constantine, as after his time the solidus became universal throughout the Empire, and has left its impress everywhere.

The crosoch therefore must be non-Roman, i.e. belong to the native population.

Above we saw that it was used along with ungas and half ungas in describing Medbh’s Fibula. Here is historical evidence of its use in the weighing of gold ornaments.

There were certainly 32 crosochs in the ounce of the Brehon Laws, but if we can show in another system of north-western Europe a weight exactly the same as the crosoch, with an ounce which is its thirty-fold, we may hesitate to lay down that the full Roman ounce with its 432 grs. Troy (576 grs. wheat) was the earliest form of Irish unga.

There is no mention of screapalls in the weight of Medbh’s brooch. It is quite possible that under ecclesiastical influences the full Roman ounce and its division into screapalls may have been introduced at a comparatively late period. The contact between Kelts and Scandinavians in early times has of late excited much interest.

2. Let us now turn to the old Norse system. It is as follows:

Let us deal first with the mark. As its name signifies, it in all probability was originally not a weight, but a measure. The use of mark as a land measure is well known in the Teutonic languages. It is also used as a measure of length. Thus a mark of cloth consists of 448 alen or ells. After what we have learned about the history of the Roman as (p. 354) we need not be surprised if a term originally used as a measure of some article which was not as yet sold by weight, came in similar fashion to be incorporated at a later period into the weight system as a higher unit. If the mark was originally a given measure of bronze or iron, we can readily see how it came later on to be used as a weight, and ultimately to be the chief unit of account among our Anglo-Saxon forefathers, until it was at last driven out by the pound.

That silver was cast into bars which weighed a mark is rendered highly probable by the fact that three of the silver bars found at Cuerdale weigh respectively 3960, 3954, and 3950 grs. Troy; that is, just the weight of 160 pennies of the reign of Alfred. 160 pennies are two-thirds of a pound of 240 pennies, or in other words a mark.

The practice of running silver into ingots of such a weight may well have arisen from an earlier practice of employing bronze or iron bars of such a weight. It is at all events certain that the mark is native Teutonic and is not borrowed from Rome. That the Kelts at least used bars of iron as money is made not unlikely by a famous passage of Caesar which I shall quote later on. A various reading states that the Britons used iron rods as money (ferreis taleis). Even without this we may reasonably infer from what we have learned of the practice of primitive peoples in dealing with iron or copper, that the Teutons and Kelts must have used these by measure. It is well known that the Swedes used ingots of copper as currency down to comparatively recent times. It is then most likely that the öre or ounce of 410 grs. was the highest original weight unit, just as the unga is in the ancient Irish system. The weight of this öre is of great interest. If we found the Roman pound of 12 ounces in Scandinavia, we should at once say that the öre of 410 grs. was the reduced Roman ounce (432 grs.). But as the native mark evidently got its position before the influence of Rome was felt in the North, we may well consider the öre to be pre-Roman. The reader will remember that I identified the ancient Roman uncia with the small talent of Sicily and Macedonia. The latter weighed 3 ox-units or about 405 grs. I also suggested that it originally represented the value of a slave, and was thus the original highest unit used for gold or silver. I showed on an earlier page (141) that the Norse örtug, the one-third of the öre, was the price of a cow. If three cows were the price of a slave in Scandinavia as they were in Ireland, and probably in Homeric Greece, an öre of gold was the price of a slave. The passage from the life of St Finnian given at once shows that an ounce of gold was the regular price of a slave in early Ireland, and probably a good Scandinavian scholar could soon find similar evidence for the value of the old Norse slave.

The meaning and derivation of the term örtug have been much discussed. It occurs in the forms örtog, örtug, ertog, œrtug. Cleasby’s Lexicon makes nothing out of the first part of the word, but takes the second part (-tog -tug = tugr = 20), because örtug had the value of 20 penningar, though tugr means 10. But as a matter of fact there were, as we saw above, 240 penningar in the mark, and therefore there were 10 penningar in the örtug. Holmboe[457] goes more deeply into the origin of örtug. He says, “As á, pl. œr, signifies a ewe, and tug-r as a derivative of ten both by itself and in compounds signifies ten, ertug seems originally to have signified 10 ewes, just as the weight ertug betokens the weight of 10 peningar, and peningr itself also means a sheep. It may be regarded as questionable to assume the plural œr to form the first part of the compound, yet œr must at an early period have been used in the formation of compounds, since both the folkspeech of Norway has the form œr-saud-ewe, sheep, technically a ewe-with-lamb, and the folkspeech of Denmark has œr lam in the sense of ewe-lamb[458].” Another suggestion is that örtug comes from arta = a pea-formed knob, so that örtug = örtu-vog, the weight of a pea.

The objection to this would be that the pea would weigh 13·5 grs. Troy, which seems far too much.

In spite of the philological difficulty in making örtug = 10 ewes, it is very remarkable that this value corresponds so accurately with the value of a cow, which I independently found for it. I have already pointed out that 10 sheep were the usual value of a cow. So it was at Rome in 451 B.C. and so it is with the Modern Ossetes. The ox fit for the yoke was probably worth 20 lambs or 5 sheep in Lusitania[459], and as we saw that in the Welsh Laws the ox when fit for the yoke was worth half a full-grown cow, the Lusitanian cow was worth 10 sheep. So also at Athens, when Plutarch[460] says an ox was worth 5 sheep, he probably means an ox fit for the yoke, the cow being worth 10 sheep. In the Brehon Laws 8 sheep go to the cow, but as I have already pointed out the insulated position of Ireland would tend to cause a variation in prices from those on the mainland of Europe. Thus we see from the story of St Finnian that gold must have been worth only three times its weight in silver in Ireland in the early centuries of our era. For the price of a slave was an ounce of gold, whilst in the Brehon Laws it is 3 ounces of silver. It might be said that we cannot prove that this was the value of a slave in gold and silver at any one time, and that silver may have been much cheaper at an earlier date. When we recollect that silver has never existed in any quantity in Ireland, and that where it does exist it can only be obtained by systematic mining, a thing impossible in the eternal turmoil of Ireland, and also bear in mind that when Japan was opened to Europeans in this century gold was exchanged for three times its weight in silver, we need not think such a relation at all unlikely in ancient Ireland. The paucity of silver ornaments in the Royal Irish Academy Museum confirms this opinion. But the evidence from the Penitentials shows that silver was scarce at a comparatively still early date in Ireland[461]. Thus XII altilia vel XIII sicli praetium unius cuiusque ancillae.

I have already shown the universality of making gold ornaments after a fixed weight. The passages given above show that a similar practice existed among the ancient Irish.

Let us turn to the numerous gold rings, commonly called Ring Money, of which there are some 50 in the Museum of the Royal Irish Academy of various weights and sizes. I give these weights. Let us examine them, and see if we can find any indications gained inductively of a weight standard.

As by inspection we see that the smallest rings weigh 13 and 14 grs. Troy, and the next three 29, 31, 32 respectively, which look like the double of the smaller, I shall group the rings according as they approximate to the multiples of 15.

A glance at the foregoing table shows that the most numerous group of rings occurs at the fourfold (60), no less than seven specimens ranging themselves at that point, next we find six specimens at the tenfold (150), whilst next in order comes the sixfold with four examples. There are three cases of the double (30). On the other hand it is worth noticing the absence of the ninefold, whilst there are three instances of the sevenfold, and the absence of the eighteenfold (2 × 9) likewise, whilst we have the elevenfold, twelvefold, thirteenfold, fourteenfold. However from the absence of the twentyfold (2 × 10) we cannot lay great stress on this. The heaviest specimen (372) closely approximates to the twenty-five fold (375).

I add the weights of the ancient Irish gold rings preserved in the British Museum.

The weights of these rings show many points of agreement with those in the Irish Museum. Thus we get 28, 29½, 30, and 32 grs. corresponding to 29, 31, and 32 grs. of the second group in the Irish Table. Again, 56 and 59½ where we get 54, 56, 58, 59, 61 in the Irish, and 66 corresponding to 65, 65; 98 to 96 and 98; and 89 corresponding to 88 and 90; 119, 121, 122 and 123 to 121 and 124; 139 to 144, and 144 and 148 to 147 and 147; then 169 to 171 and 172. Then comes a break, and we get 215, 218, 218, 224 corresponding to 220, and 249 to 247, and 283 to 283 and 283; and 323 to 322, and 360 to 366. But the British Museum gives us in the higher weights three very important specimens: for 410 grs. is the ounce corresponding exactly to the old Norse öre of 410 grs., and the ring of 422 grs. looks like the later ounce rising towards the full weight of 432. The ring of 806 grs. is plainly 2 ounces of the standard of 410 (806 ÷ 2 = 403).

The occurrence of several specimens so constantly all of the same weight, as for instance those about 220 grs., points beyond doubt to the conclusion that when the rings were being made a given quantity of gold was weighed out for the purpose. The story of St Finnian proves that for any transaction in which rings were employed as money, the scales were employed.

There is a set of leaden weights in the Royal Irish Academy Collection, found at Island Bridge, Dublin, in 1869, when Ancient Irish and Scandinavian remains were found together. As they are more or less corroded, it is not advisable to lay much stress on their present weights.

There are certainly some interesting points of agreement between the weights and the gold ornaments, e.g. the weights of 220, 390, 414, 630, have corresponding weights in gold. The largest weight may be 4½ oz. of 410 grs.

Let us now return to the Irish monetary system, and see if we can determine more accurately its relation to that of Rome.

As regards unga and screapall we have spoken already. Of their origin there is no doubt. The pinginn on the other hand is not so easy. The name is certainly Teutonic, said to be ultimately a loan word formed from pecunia. It seems to have been employed as a general term for the smallest form of currency. Hence we find the Saxon form (pendinga) applied to the 240th part of the lb., and of about 32 grs. wheat, and the Norse peningr used for the 240th part of the mark, whilst in Ireland the cognate form is applied to the 72nd part of the ounce, and is of the weight of 8 grains wheat.

The Irish employed the system of Uncia and Scripula. Shall we say then that this system was in vogue in Britain likewise before the time of Constantine and yielded slowly before the later one?

Since then it was common to the Kelts on both sides of the Irish Sea, and we find that in Ireland it was grafted upon an earlier system, of which the crosoch is a survival, we may reasonably infer that the Kelts of Britain had likewise a native system analogous to the crosoch. But further, of this we have strong evidence of two kinds. Caesar B. G. v. 12, when describing the British Kelts and their manners, says; pecorum magnus numerus. Utuntur aut aere aut nummo aureo aut annulis ferreis ad certum pondus examinatis pro nummo[462]. The passage has been mutilated by Editors, but this is the reading of the best MSS. Caesar thus tells us that they had a system of weights of their own. Secondly the evidence of the actual British Coins (cf. Evans, Coins of Ancient Britons) which are of a standard not Roman.

Now we have seen above that the Irish gold rings were weighed on a standard of almost 13·5 grs. Troy. Let us now see if the larger gold ornaments preserved in our Museums confirm or disprove the evidence of the rings. I shall first give the weights of those in the Royal Irish Academy[463]:

The result of an examination of the foregoing weights is to show that in all probability the vast majority of them were made on a standard much lighter than the Roman ounce of 432 grs., which was in full use in mediaeval Ireland. We saw that the Roman ounce had been only 420 grs. down to the Second Punic war, and I suggested that originally it was of the same weight as the Sicilian talent 390-405 grs. Can we observe a similar increase in the Irish ounce? The ounce of 400-410 seems to point to a time when Kelt and Scandinavian had a common higher unit of similar weight corresponding to the value of a slave[464], just as the Sicilian and Macedonian talent of three ox units represented the same slave unit.

I shall now give the weights of the various ornaments of gold found in England, Wales and Scotland which are preserved in the British Museum. For these I am indebted to the great kindness of Mr F. L. Griffith of the Anthropological department.

The evidence points to an ounce of 420 grs. It is worth noting that this is just 5 times the weight of the latest British coins, 84-82 grs.

Whence then did the Britons obtain this pre-Roman standard? Was it of native development or borrowed from some other people? By Britons we must be careful to express not all the natives of Britain. They fall most certainly into at least two groups. I. The Kelts in the East and South East. II. The barbarous inhabitants of the interior, who subsisted by hunting and fishing, and who were probably of that Iberic race, which spread over all Western Europe before the advance of the Aryans. It is only with the first group that we are immediately concerned. They almost exclusively possessed the art of coining, as is shown by the area over which British coins are found. Furthermore Caesar tells us of the close relationship of the first group to the Gauls, as is shown by their tribal names, language and customs. In addition their coinage is similar. Now there can be no doubt as regards the source from whence the Gauls derived their coinage. As they got the art of writing from the Phocaeans of Massilia (founded circ. 600 B.C.), so likewise did they gain the art of money-stamping from the same famous town, as has been completely demonstrated long since. People are inclined at once to assume that the Gauls and Britons got their weight standards also from Marseilles. There is certainly some evidence to support this belief. Thus the gold torque lately found in Jersey weighs 11500 grs., which is exactly the mina of the Phocaic system at a time when 57½ grs. went to the drachm. Again we have seen that there were a considerable number of gold ornaments in Ireland and Britain which weigh 224-216 grs. This is the Phocaic (or Phoenician) stater. But the question is not so simple as it might appear at first sight in relation to the weight system, as will appear most readily by a short survey of the history of the monetary system of Massilia.

I. The earliest coinage consists of silver, small divisions of the Phocaic drachm (58-54 grains Troy). These have various symbols on the obverse, but have uniformly the incuse square on the reverse. These may be placed after 500 B.C. “Notwithstanding their archaic appearance, it does not seem that these little coins are much earlier than the middle of the 5th century.”

II. Next comes a series, chiefly obols for the most part with head of Apollo on obverse, and a wheel on reverse, the latter probably a development of the earlier incuse square. They are mostly obols of 13-8 grains.

III. About the middle of the 4th century the drachm first appears with the head of Artemis on obverse and a lion on the reverse, weighing 58-55 grains.

Now over all Gaul, and far into Northern Italy, and the valleys of the Alps, as far as the Tyrol, the coinage of Massilia made its way and was abundantly imitated. In fact these imitations formed the entire medium of those regions until the Roman conquest. The imitations of the little coins with Apollo and the wheel as reverse are found right into the north of France, and in England.

Did the Kelts borrow their 13½ grain unit from the 13 grain obol of Massilia, or is it of far earlier growth? The Etruscans used a unit of 13½ grs. in the 4th century B.C., and we find the Massaliotes having almost the same. Is the true answer this? All over Western Europe the ox unit of 135 grs. of gold was subdivided into 10 parts each of 13½ grs. These 10 parts corresponded to 10 sheep, the regular value of a cow. There was also a higher unit from Greece to Gaul and Britain corresponding to the slave. There were fluctuations in their worth in various times and places, but on the whole there was a tendency to raise the weight of the higher unit (ounce). But it is natural that the Kelts may have taken over into their system certain units from the Phocaic system which they used as multiples of their own smaller units, just as the Teutonic peoples took the Roman pound into their own system, and the natives of West Africa made the Spanish dollar the multiple of their own native weights, based on seeds. Some idea of the relative ages of Keltic gold ornaments may perhaps be got from applying the criterion of weight standard to them.