Up to the time of the Revolution each province had its own measures and weights, more or less influenced by the uncertain standard measures of the king in Paris. This was the effect of the feudal system and of the very gradual annexation of the provinces under conditions which left considerable powers to the parliaments and other local authorities. Even in each province varieties of measures were to be found, and they exist to this day in each pays, often in each parish.

The basis of this very loose system was Roman, influenced in the North by Teutonic importations, but especially by the peculiar and intrinsically perfect system of the South, where the Roman basis had entirely disappeared under the influences of commerce with Egypt and with that portion of Africa which begins across the Pyrenees, and which in medieval times imparted much of its high civilisation to European countries.

1. The Southern System

This system, prevailing far beyond the limits of Occitania, the land of the Lengo d’O, had for its basis the Load of Wheat, a measure very nearly that of the cubed Arabic cubit, and comparable with the English Coomb or half-Quarter. Just as the English Quarter of corn is 8 bushels, so the Cargo (load, or Saumado, ass-load, Seam) is 8 Eimino. And just as we had a wine-bushel, originally a cubic foot in water-wheat ratio with the corn-bushel, so Occitania had its Escandau for wine corresponding, in the Southern water-wheat ratio of 1 to 1·22, to the Eimino or Panau. The only difference in this evolution was that, while our corn-measure was increased from the wine-measure, the southern wine-measure, and other measures evolved inversely from it, were produced from the corn-measure as a basis. The word Escandau means ‘standard’ (like the Denerel of Guernsey), and just as the cubic measure, the quadrantal, of 1000 Roman ounces of water, is the standard of our foot and virtually of all our other measures, so the Escandau-quadrantal is the standard of the PÁn and of all the other measures of Marseilles. I take the standards of Marseilles as it was the great port of trade in the South, and incidentally those of Arles, the capital of the medieval kingdom of Arles or of Burgundy, afterwards the republic of Arles. This was so considerable a seaport, connected as it was with the sea both by the Rhone and by a canal passage, the FossÆ MarianÆ, through the lagoons, that at one time the Lion of Arles was a rival of its brother of St. Mark, and gave its name to the Gulf which receives the Rhone.

The process of involution by which the PÁn of Marseilles was derived from the side of an Escandau of quadrantal form has been described in Chapter IV.

The Cano or fathom, = 79·24 inches, was 8 pÁn or spans each = 9·904 inches; the span was of 8 menut or inches, also divided into 8 parts.^[47]

The basis of the Southern system, typically that of Marseilles, was then the Cargo, a corn-measure = 34·73 gallons (the equivalent of 154·79 litres, the official metric value), which was the cubic cubit of Al-Mamun:

Now what water or wine measure would be produced from the Cargo, decreased in wheat-water ratio?

Dividing the measure of the cargo by 1·22 we have:

A fluid measure of this capacity is not in use at Marseilles, but we find its half, almost exactly, in the Mieirolo = 14·19 gallons, a wine and oil measure used extensively in Mediterranean ports.

The word Mieirolo, in which miÉ means half, corresponds to the name of the first in an Italian series of wine-measures:

Mezzaruola, Terzaruola, Quartaruola, fractions of a 28-gallon measure now apparently obsolete.

The standard of the Mieirolo is now at—

One-fourth of the Mieirolo, or one-eighth of the obsolete wine-cargo, is the Escandau, equal to the Spanish arroba (a word meaning ‘quarter’), and containing, at the present Marseilles standard, 16·096 litres = 3·54 gallons. To this Escandau or standard corresponds, in water-wheat ratio, the Panau = 4·34 gallons, 1/8 of the Cargo = 4·34 bushels or 34·73 gallons.

The correspondence of this series of wine and corn measures, in southern water-wheat ratio, is perfect, even after many centuries, probably since the tenth century. The Escandau and the Panau or Eimino correspond then to about 4 wine-gallons and 4 corn-gallons.

The Escandau has always been understood to be a cubic pÁn. Escandau^[48] means a standard; PÁn means a side, pane or panel, and it is the measure of the side of a ‘quadrantal’ containing an Escandau of water, as our foot is the measure of one containing an English talent of 1000 Roman ounces of water. The cube root of 16·096 litres is 25·24 centimetres, a length differing by less than a millimetre from the standard of the Marseilles pÁn = 25·16 centimetres or 9·9 inches.

Land-measures

The ancient system of seed-measures, fixed geometrically, survives to this day in Southern France, indeed throughout most of France. I shall make no apology for dwelling on it, for the linear land and cubic measures of Southern France show a perfectly concordant system of measures, more so even than those of England; indeed they are the type of a perfect system.

The largest unit of land is the Saumado, of 4 Sesteirado, each of 2 Eiminado; these being originally the ground that could be sown with a Saumado (or Cargo), with a SestiÉ, with an Eimino, of wheat.

These seed-measures of land corresponding to our Coomb, Bushel and Peck land, became fixed respectively at 1600, at 400, and at 200 square cano or fathoms.

To the SestiÉ and the Sesteirado correspond the boisseau and boisselÉe of Poitou and other provinces, the boisselÉe, or bushel-land, being 400 square toises.

But the surveyor’s measuring-rod is the Destre, a double cano, of 16 pÁn = 13 ft. 2-1/2 in. In Languedoc, west of the Rhone, the square destre = 4 square cano is the smallest unit, so that the Saumado of land is 1600 square cano or 400 destre. But in Provence the destre of land is 2 square cano, so that the Saumado is 1600 square cano or 800 destre; the reason probably being that the destre should be 2 cano superficial as it is 2 cano linear, and also that the Eiminado or peck-seedlip of land should be 100 destre.

The Eiminado is divided into quarters and sixteenths, corresponding to the gallon and quart divisions of the Eimino or peck. It is also divided into 20 Cosso, the ground corresponding to a cosso (= quart, wine-measure) of seed.

It is interesting to observe that the Saumado of 4 Sesteirado of 40 Cosso, corresponds, in division, to our Acre of 4 roods, of 40 square rods.^[49] And the Cosso = 1/100 acre or 1/10 sq. chain.

N.B.—1000 sq. cano = 1 acre.

The Saumado, of 1600 sq. cano = 1·6 acre.

Such is the typical system of Southern measures, best preserved in the neighbourhood of Marseilles, but prevailing throughout the Southern half of France, though with local variations in the length of the cano and the names of the land-units.

Measures of Capacity

These have mostly been given in the story of the pÁn and in the seed-measures corresponding to the land-measures.

Throughout the system the divisions in each series are sexdecimal, even the Cosso, 1/20 Eiminado, being 1/160 Saumado.

Weights

There were three types of pounds in South France, local variations from these being very slight. The pound was always 16 ounces, each of 8 ternau. The Ternau, so called from its being divided into 3 pennyweights, was the Arab dirhem. The three types of pound were:

The Quintal was 100 of these pounds, but long hundredweights were common. Its quarter was the Rub (Ar. rouba, four). These weights are nearly obsolete, as the possession of any weights not of the Republican system would be illegal. The measures of length and capacity are often slightly altered so as to be in metric units: the pÁn becomes a quarter-metre; groceries are often ticketed by the hectogramme, as this is known to coincide very closely with the old Southern quarter-pound.

We now pass to the Northern or Paris system, mostly taken from the South, and bearing evident traces of this origin.

2. The Northern System

Measures of Length

The Roman foot survived in North France as the quarter of the Aune or ell, a measure = 46·77 inches. (Cf. the passetto or double braccio of Tuscany, of 4 palmi = 45·96 inches.) As a cloth-measure the Aune was divided, like our cloth-yard and ell, into eighths and sixteenths.

But there was also the pied de roi, the royal foot, one-sixth of the Toise, which = 76·73 inches = 1·949 metre.

The royal foot, = 12·789 inches, was divided into inches (pouces) of 12 lines, each of 12 points. Its standard was traditionally referred to Charlemagne, either to the length of his foot, or to a standard brought to him by the envoys of Harun-al-Rashid. It coincides with half a HashÍmi cubit, 25·56/2 = 12·78 inches. This tradition must be dismissed; new measures are not introduced as standards in that way. It was simply one-sixth of the toise, which was a Cano from civilised South France, but its standard was so ill-kept as to be of doubtful exactitude. All that is known of its standard is that, about 1668, an iron rod was fixed in a wall of the Grand Chatelet in Paris and that the length of this rod was that of half the breadth of the eastern gateway of the Louvre-palace, which gateway was, according to the plans, 12 feet in breadth. This standard was, however, considered to be 5/12 inch short of the customary toise.

The Louvre standard, taking it at = 1·965 metre (which I find it by actual measurement), corresponds closely to the Cano of Beaucaire. This town on the southern Rhone, opposite Tarascon, had a great annual fair, and may thus have given its linear standard to trade in the same way that Marseilles passed the Cargo of its Egyptian corn-trade on to Paris as the Setier, and that Troyes passed the marc used at its great annual fair on to Paris as the standard of the French troy pound.^[50]

But the royal foot was inconveniently long for popular use, and a practice arose of taking 11 inches of it as a customary foot = 11·7 inches. This reduced foot, coinciding almost exactly with the quarter-Aune, was much used in the districts north of Paris as the pied de Ponthieu, or de Clermont. The Brasse was a short fathom of 5 pieds = 5 ft. 4 in., probably an adaptation of the Roman pace. A pas (pace), of half a brasse = 32 inches, is used in some districts for land-measurement.

Measures of Distance

There was no official measure of distance, such as our furlong and mile, between the toise and the league, and the league was very variable (see Chapter III). Acre-lengths, cordes, and other popular measures supplied the want, more or less well. In some districts (also in Mauritius) there were milestones at intervals of 1000 toises, called a mille. In South France the mille was divided into centeniÉ of 100 toises or perhaps local cano. This was probably the length of the sesteirado, the rood, 100 × 4 cano.

The corde, a field-measure used before the surveyor’s chain, was of variable length. In Burgundy the league of 3000 toises was divided for roadwork into 50 portÉes, of 12 cordes; these would thus be 5 toises or 30 feet. But there seems also to have been a corde of 33 feet, perhaps reduced feet, and thus = 30 royal feet, and this, doubled, was used as the rough measure of a ‘cord’ of firewood = 4 × 4 feet, in 4-foot logs. This is the probable origin of our ‘cord-wood’ as applied to stacked logs for fuel.

The units are the square toise = 4·543 sq. yards, the perche and the arpent, with other units in local usage.

There were three different perches officially recognised, and still in common use.

1. Perche d’ordonnance or of the Eaux et ForÊts administration, 22 royal feet = 23·466 English feet; the square perch of 484 sq. feet = 13·44 sq. toises = 2 sq. rods.

The approximate coincidence of the quarter-aune with the reduced royal foot, i.e. of 12 Roman inches with 11 royal inches, was the probable reason of the standard perch being fixed at 22 feet = 24 Roman feet or 6 aunes.

The standard arpent was 100 square perches = 1344 sq. toises = 200 rods or 1·26 acre.

2. Perche commune, 20 royal feet = 21·3 English feet, the square perch of 400 sq. feet = 11·11 sq. toises = 50·47 sq. yards.

The arpent commun was 100 of these square perches = 1111 sq. toises = 1·04 acre.

3. Perche de Paris, 18 royal feet = 19·83 English feet, the square perch of 324 sq. feet = 9 sq. toises = 40·9 sq. yards.

The arpent de Paris was 100 of these square perches = 900 sq. toises = 0·844 acre.

The arpent commun is that of Quebec.

The arpent de Paris is that of Mauritius.

The acre de Normandie varies according to its perch, but it is always 160 sq. perches, and if these be standard it is equal to 2 acres. But the usual unit is the vergÉe or rood, of 40 perches = 1/2 acre.

It has been seen that the Jersey vergÉe is 40 perches of 22 reduced English feet square, the foot being 11 inches. This is an adaptation of a very general Normandy perch, 22 feet of 11 French inches. It is = 0·44 acre.

Local French land-measures varied considerably, from different standards of perch, from different lengths taken for the foot of the perch. But the size of the unit, Journal, EstrÉe, &c., &c., is very generally = 1400 to 1600 square perches or roughly about 1-1/2 acre. These measures, so irrational to the Parisian, are dear to the peasant’s heart; he understands them, and as people do not buy land as they would apples or eggs, no one is deceived.

The EstrÉe or SeterÉe (Setier seed-land) might be divided into 12 BoisselÉes (small-bushel lands).

Weight

The royal pound, livre poids de marc, the double-marc of Troyes, was one of several pounds current in Northern France. It was, like the royal foot, ascribed to Charlemagne, but his standard of weight, as known by his silver pennies, nearly always much above 24 grains, 1/20 of some ounce heavier than that of the Troyes marc, was probably altered later on. The royal pound, = 5570 grains, was raised for commercial purposes (about 1350) to 16 ounces = 7554·1 grains, the ounce = 472·13 grains.

The weight of the 12-ounce pound coincides very closely with that of the Bosphoric minÁ, 100 drachmÆ of 56·66 grains; this is perhaps the origin of the story that it was sent to Charlemagne by Harun al Rashid. Its ounce is also approximately the Tripoli ukyÉ of 10 dirhems × 470-3/4 grains, and nearer still to 471 grains, the weight of 10 of the dirhems of which 8 made the ProvenÇal ounce.

It is probable that the French pound was one of the lighter pounds of the variable Northern Troy series, all with an ounce of 10 dirhems of 48 grains more or less.

The ounce was divided into 8 gros, groats or drachms, of 3 deniers or dwt., each of 24 grains. So the livre was 16 × 24 × 24 = 9216 French grains. These were light grains, not the heavy grains, 20 × 24 to the ounce, of English and other mint-weights.

There was a Quintal of 100 livres = 107·7 lb.

The Tonne or tonneau was 2000 livres = 2154 lb.

Value

The French coinage-system, probably instituted by Charlemagne, was the same as ours. The original unit was the silver penny, estelin (sterling) or denier (L. denarius) of 24 French grains; 12 deniers made a sol or sou (L. solidus, shilling) and 20 sols made the livre or pound, originally a livre d’estelins, a 12-ounce pound of sterlings. But the silver coinage shrank and was debased, until, by the eighteenth century, the pound, livre or franc was a silver coin worth tenpence, the sol a copper halfpenny, and the denier had shrunk, even as copper, to so minute a size that its place was taken by the liard, a small copper coin of 3 deniers, a quarter-sou; even the double of 2 deniers had disappeared. Accounts were kept in livres and sols and deniers, our £ s. d., but at 1/25 the present value of our coin.

The Écu of 3 livres, that is of 60 sous, was largely used; wages of farm-servants are often at the present day reckoned in Écus. This was properly a petit-Écu or half-crown, but the real Écu of 6 livres was so little used that the smaller coin took its name. And, as our half-crown has the great convenience of being one-eighth of a sovereign, so the Écu had that of being one-eighth of a louis, the gold piece of 24 livres. This was the value of the louis at par, for it varied as did that of the guinea when England was a silver-standard country.

Measures of Capacity

These measures, both the wine-series and the corn-series, were quite discordant and had no relation to the measures of length. That this was caused by an incoherent system of factors is shown by there being in each series a unit derived from the perfectly concordant measures of the South:

The wine-velte = 1·76 gallon, half of the Escandau.

The corn-setier = 34·32 gallons, the Marseilles Cargo.

The former, when increased in water-wheat ratio, is almost exactly 1/16 of the latter. So, had the former, increased in this ratio, been multiplied sexdecimally, concordance would have been preserved. But there was a customary Muid = 63-1/2 gallons, our hogshead, with its quarter, our kilderkin, the Quartaut = 15·8 gallons, and not to derange these measures the velte was made one-ninth of the Quartaut. And in the corn-series the Setier was divided and multiplied duodecimally. So the concordance was entirely deranged.

1. Wine-measures.—The Velte (the origin of which is given in Chapter XIX) was divided into 2 gallons (our wine-gallon), 4 pots (our pottle), 8 pintes. The last of these, = 1·76 pint, was about our old wine-quart, = 32 oz., its half was a chopine or setier, = our wine-pint, and the half of this was the demi-setier, a name still current, the French equivalent of our popular ‘half-pint.’

2. Corn-measures.—The standard unit was the Setier = 34·32 gallons, or 4·29 bushels, differing very slightly from the Marseilles Cargo = 4·34 bushels. As the Setier was an isolated measure, while the Cargo was from early medieval times the basis of the complete system of Southern measures, it may confidently be inferred that the Paris unit of corn-measure was taken from that of Marseilles, which was the Egyptian Rebekeh, the cubed Arabic cubit.

The term Setier is the L. sextuarius, but it had lost its original meaning and become a general-utility term in measures. The Setier = the Marseilles Cargo of 4 SestiÉ, must not be confused with this sestiÉ. It was divided into 12 boisseaux of variable standard, but usually estimated to hold 20 French pounds of wheat. As 1/12 setier, the boisseau was = 2·86 gallons, and it was divided into 16 litrons = 1·43 pint.

There were intermediate divisions of the Setier; it was of 2 mines (a term taken from the Southern eimino), 4 minots, 12 boisseaux.

There was also a Muid for corn and salt. The corn-muid was 12 setiers.

There are still in France traces of an older system of corn-measures derived from the cubic foot. I found, in the Rouen Museum, the standard bushel of the town of Bolbec. It measures 16 inches diameter by 12·6 inches deep = 2533 cubic inches or 9·14 gallons. It appears to be the French cubic foot = 2091 cubic inches increased in water-wheat ratio to 2533 × 1·22 = 2551 cubic inches, a difference probably to be ascribed to the difficulty in measuring at all accurately.

There are also many local standards of capacity, well deserving of study. Some, as the bushel of La Rochelle, indeed of the west of France generally, = 56 lb. of wheat, are much larger than the Paris Bushel. There was a general rejection of the duodecimal division of the Setier.

Remarks on the French Measures of Capacity

The fault of the Paris system was that there was little or no concordance between the different series.

In length, 6 aunes approximately coincided with 22 feet or 3-2/3 toises.

The measures of length had no concordance with those of capacity, and in the latter, wine-measure and corn-measure had lost their original concordance when they were brought from the south. They lost it by two faults:

1. By making the quartaut of 9 veltes instead of 8;

2. By dividing the setier into 12 boisseaux instead of 8.

Had this octonary division been substituted, it would have been quite satisfactory, and concordance with the linear standard would have been obtained.

A quartaut of 8 veltes, 8 × 1·76 = 14·08 gallons, would have been in water-wheat ratio with the corn half-setier = 17·16 gallons:

And the setier divided into 8 parts would have given a larger boisseau = 4·29 gallons (a peck) corresponding in water-wheat ratio to the double velte of 4 gallons and measuring approximately 1000 cubic pouces (983 exactly); its side, when of cubic form, being almost 10 pouces, and thus affording an easily applied linear measurement as a check on the variation of the boisseau. The standard of this measure was most variable from want of such a check. Really, as 1/12 Setier it should have been 655·4 cubic pouces, but it varied between 644 and 677, its reputed capacity being 640 cubic pouces.

It would have been easy to have fixed the new boisseau at 1000 cubic pouces, raising the variable standard of the Setier to 8000 cubic pouces = 34·9 gallons instead of its reputed standard = 34·32 gallons.

By these slight alterations perfect accordance with the southern measures would also have been obtained.

Leaving the measures of length and surface which were sufficiently concordant, the measures of capacity would have been:

A water-wheat ratio of 1:1·24 would have been preserved between the two series, and their connection with linear measures through a cubic boisseau of 10 pouces each side (or a cylindrical one of 10 pouces diameter and 11·4 pouces in height) would have been most advantageous.

It may seem futile to make these proportions 120 years too late, but they may be useful in showing how unnecessary was the revolutionary plan of uprooting the old measures.