Ancient Agriculture

by G. W. B. HUNTINGFORD

I N the first volume of ANTIQUITY appear two papers, one by Dr

R. C. C. Clay, dealing with the formation of lynchetsl ; the other,
by Dr E. Cecil Curwen, containing a survey of prehistoric agriculture
in Britain.2 These papers, which are of considerable interest to the
farmer as well as to the archaeologist, have suggested the following
remarks, which I was unable to put on paper before as some of my
books were in England.
It may be as well to note first of all the statements on which
this paper will comment. Dr Curwen says that Early Iron Age and
Romano-British ploughs had ploughshares consisting of a plain metal
point fitted to the share-beam-such a plough merely scratches a groove
(p. 268) ; lynchets were not formed intentionally (p. 273) ; a plough
described by Pliny which turned a furrow did away with the need of
cross-ploughing (p. 280) ; manuring is implied by rectangular fields
with lynchets, and no visible evidence of two- or three-field rotation of
crops (p. 286). In Dr Clay’s paper, lynchets are defined as ploughed
ground of which the natural slope has been altered by ploughing, which
alteration was effected by ploughing the lower furrow only, bringing
the plough back idle. This mode of ploughing flattened the slope, a
slope being unsuitable for growing corn in a damp climate-the climate
of Britain being damper in prehistoric times than it is now. (Pp. 57-59).
I. THEPLOUGH
The essential feature of the plough is that it turns a furrow by
inverting the soil, whereas the cultivator or horse-hoe for working after
the crop is sown merely stirs the soil without inverting it. Dr Curwen
describes Early Iron Age and Romano-British ploughshares as con-
sisting of a simple metal point3 which, he thinks, merely scratched a
groove in the soil (pp. 268, 280). T o this failure to turn over the
soil he considers that we are to ascribe the need for cross-ploughing
‘ Some Prehistoric Ways ’, ANTIQUITY, 1927, I, 54-65.
2 ‘ Prehistoric Agriculture in Britain ’, Ib., I, 261-89.
3 Our ‘ bar-point ’ share.

327
 ANTIQUITY
advocated by Roman agricultural writer^.^ This, as I shall endeavour to
show, is not so. Further, I believe Dr Curwen to be mistaken in his
idea of the ancient plough ; for, among other things, practical experience
has shown me that no amount of cultivating, i.e. stirring the soil without
inverting it, will produce a lynchet or terrace.
Till 1760, practically the whole of the plough was universally
made of wood, and there is reason to believe that the shape of the plough
changed but little during centuries of use.6 The most detailed account
we possess of a Roman plough is Vergil’s, which runs as follows :-
Continuo in silvis magna vi flexa domatur
In burim, et curvi formam accipit ulmus aratri.
Huic ab stirpe pedes temo protentus in octo,
Binae aures, duplici aptantur dentalia dorso.
Caeditur et tilia ante iugo levis, altaque fagus,
Stivaque, quae currus a tergo torqueat imos :
Et suspensa focis explorat robora fumus.
[In the first place, an elm is bent forcibly in the woods to the shape
of a plough-body (buris), and takes the form of the crooked plough.
From the base of this extends the beam (temo) for eight feet, and joined
to it are two mould-boards (aures) and slades (dentalia) with double
back. Beforehand, too, is cut the light lime for the yoke, and the tall
beach for the stilt (stiva), which turns the bottom of the fore-carriage
) ~ behind : the wood is hung in the chimney to be seasoned
( c u r r u ~ from
by the smoke].’
To illustrate this plough, I append tracings from Martyns of two
Italian ploughs in use in the 18th century, with another drawing of an
old plough formerly used in Sussex. (Figs. I, 2, 3).
Now, as regards the work done by the Roman plough, we have
Vergil’s express statement that
Pingue solium primis extemplo a mensibus anni
Fortes invertant tauri, glaebasque iacentis
Pulverulenta coquat maturis solibus a e s t a ~ . ~
-~

4e.g. Vergil; Varro, R.R. I, XXIX, 2 ; and cf. Festus, S.V. Offringi, p. 523 (ed.
Dacier, Valpy’s ed., London, 1826).
Fream, Elements of Agriculture, ed. 10, p. 47. Fream is also mistaken, I think,
in supposing that the old ploughs did not invert the soil.
6 I understand ‘ currus ’ as a wheeled forecarriage on the strength of Servius’
comment. See Conington in loc; see also Pliny, XVIII, §18,48.
Verg., Georg., I, 169-175.
‘ The Georgicks of Virgil ’, by John Martyn, F.R.s., ed. 5 (Oxford 1827), p. 40.
Figs. I, 2, 3.
Verg., Georg., I, 64.
328
 H

FIO.I. EIGHTEENTH-CENTURY ITALIAN PLOUGH. (Martyn)

A, Buris. B, Stiva. C, Temo. D, Aures. E, Dentale. F, Vomer. G, Culter. H, Tabellae

FIG.2. EIGHTEENTH-CENTURY MANTUAN PLOUGH. (Martyn)

A, Buris. B, Stiva. C, Temo. D, Aures and dentale. E, Vomer. F, Culter. G, Currus

FIG.3. OLD SUSSEX PLOUGH. (Reliquary)

329
 ANTIQUITY
(Right from the beginning of the year the strong bullocks should turn
over the rich soil, that dusty summer with its hot suns may bake the
clods as they lie [upturned]). We learn from other sources something
of the nature of ploughed furrows. ' Qua aratrum vomere lacunam
striam fecit, sulcus vocatur. Quod est inter duos sulcos elata terra
dicitur porca, quod ea seges frumentum porricit 'Jo w h e r e the plough
makes with its share a hollow furrow, the result is called sulcus (furrow).
The raised earth between two furrows is called porca (ridge), because
there the corn produces its crop]. And again, ' Porca est inter duos
sulcos terra eminens',11 (porca is the raised ground between two
furrows) ; which is further defined by Columella, ' Liras rustici vocant
easdem porcas cum sic aratum est ut inter duos latius distantes sulcos

B B B

FIG.4. DIAGRAM OF FURROWS. AA, Sulcus ; BB, Porcae

medius cumulus siccam sedem frumentis praebeat ',l2 (the country

people call these same ' porcae ' lirae when the land is so ploughed that
a heap of earth mid-way between two wide furrows offers a dry bed
for the crop). This result (fig. 4), can only be obtained by using a
plough which turns a furrow. Such a seed-bed was produced, so
Varro tells us, ' cum tabellis additis ad vomerem ',I3 (with boards added
to the share), i.e. not to the actual share, but to the mould-boards
proper (fig. I ) . The use of a coulter, too, even if not with every
plough, implies the inversion of the soil, particularly as Pliny says
' culter vocatur, praedensam, priusquam proscindatur, terram secans,
futurisque sulcis vestigia praescribens incisuris, quas resupinus14 in

1°Varro, R.R., I, 29, 43. Ed. Goetz (Teubner).

Festus, p. 319, S.V. Imporcitor. The porcae described above must not be con-
fused with others 'aquae derivandae gratia' (ib. p. ~ I I )which
, were open drains.
l2 R.R. 11, 4, 8.
l3R.R. I, 29, 92.

l4Taking 'resupinus ' proleptically.

330
 ANCIENT AGRICULTURE
arando mordeat vomer 6:’ (that which cuts the heavy ground before
the first ploughing is called a coulter, marking out a line for the future
furrows by cutting into the ground, into which cutting the share
penetrates turned back in ploughing). There is no other reasonable
explanation of the term porca than that shown in fig. 4 ; and the state-
ments of Varro, Vergil, Columella, Pliny, and Festus sufficiently
indicate that the ground was actually turned over, not merely scratched.
The only alternative is to suppose that every second furrow was left
unploughed, which is absurd, and could not possibly have prompted
Varro’s etymology of porca, ‘ quod ea seges frumentum porricit ’ ;le
nor does this sanction White and Riddle’s definition of porca as a
‘ balk ’.
Dr Curwen considers (p. 280) that a plough which turns a furrow
does away with the need for cross-ploughing. But, apart from the fact
that the Roman plough did turn a furrow, we need not suppose that
the terms proscindere =first ploughing, offringere=second or cross-
ploughing (‘ to plough against ’), and lirare=third ploughing (after or
at the same time as sowing), need be taken at any period of Roman
agriculture otherwise than literally : husbandry is the most conservative
of all arts. And, though these methods are not now practised in England,
the same procedure is adopted in maize-growing countries. Thus, in
Kenya, our normal procedure is, where conditions permit, to plough
and cross-plough, with possibly a third ploughing; the seed, too, is
often planted behind the plough, and immediately harrowed in. And
our ploughs are infinitely superior to anything the Romans devised.
When the plough is drawn by oxen (whether the team be of two or
fourteen) a cross-ploughing is often necessary because an ox-team does
not plough as straight a furrow as a horse-team, and there are generally
places which the plough has missed in the first ploughing. Further,
the damper the climate, the worse the weeds, and one ploughing is often
not enough to cover the weeds properly. Therefore, I maintain that the
type of ox-drawn plough17 used-provided it turns a furrow-cannot
be said to affect the need for cross-ploughing.
‘ Ploughshares of the Early Iron Age and Roman period are not
l5 N.H. XVIII, $18, 48. And Pliny’s words ‘Latitudo vomeris cespites versat’ (loc.
cit.). are conclusive.
l6 The probable etymology of ‘ porca’ is porcus, and the literal meaning ‘ little pig ’.
1’ Only when a tractor is used is one ploughing sufficient under normal tropical
conditions.

331
 ANTIQUITY
uncommon, and consist of a simple metal point designed to fit on to
the share-beam, without any device for undercutting and turning
over the sods. Such a plough simply scratches a groove in the soil’.
(Cunven, p. 268). But the discovery of metal shares only does not
necessarily mean that the ploughs they belonged to did not possess
other parts made of wood which could turn over the sods. The Roman
plough was made of wood, and wooden mould-boards were used in
Britain as late as 1 8 3 0 ; ~ ~ in the Sussex plough already referred to
and
even the coulter is made of wood.10 Hence we may reasonably infer
that people who had the intelligence to make an iron share, had also
sufficient intelligence to provide their ploughs with some means of
inverting the soil ; for inversion is absolutely necessary in levelling a
slope ; and we may conclude that the mould-boards, together with the
rest of the ploughs, have perished. A drawing in a MS of Caedmon’s
Paraphrase shows that the ‘ Vergilian ’ or Italian type of plough was
used in England from early modern times ; this plough has a mould-
board. (Fig. 5 ) .
When we approach the problem of ploughs of earlier periods than
the Iron Age, we are on far less sure ground. That ploughs of a sort
were used is quite clear ; not so clear, whether they ploughed a furrow
or scratched a groove.
The ploughs figured in the rock-carvings from the Maritime Alpsz0
appear to possess a wooden bar which was driven through the ground
(fig. 6)’ like the ancient Egyptian plough (fig. 7)21 which, we know,
merely scratched a groove in the soil. On some of the megalithic
remains near Carnac in Brittany are sculptured figures which M. le
Rouzic believes to be ploughs (‘ hache-charrue ’). Examples from the
Dolmen des Marchands and the tumulus of Mant-er-H’roek at
Locmariaker are shown in figs. 8 and 9 . 2 2 These things, if they do
represent ploughs, can have done no better work than the Alpine and
Egyptian implements. My contention, however, that the ploughs which
worked in the Early Iron Age fields had mould-boards, does not apply
to ploughs of the Neolithic, Copper and Bronze Ages.
l8 Fream, loc. cit., 47.
Reliquary, N.S. X I , 219.
2o M. C. Burkitt, Our Earb Ancestors, plate 28, fig. I .
21 After Wilkinson, Ancient Egyptians, abridged ed., 1854,11, fig. 359.
22 2. le Rouzic and C, Keller, Locmariaker : La Table des Marchands. (Nancy,
1910). My own copies of these are at present inaccessible to me ; figs. 8 and 9 are
therefore after Le Rouzic.
332
 ANCIENT AGRICULTURE

FIG.5 . ' VERGILIAN ' TYPE OF PLOUGH. (Caedmon's ' Paraphrase '1

FIG.6. PREHISTORIC PLOUGH AND HARROW. (Burkitt)

By permission, Cambridge University Press

FIG.7. ANCIENT EGYPTIAN PLOUGH. (Wilkion)

333
 ANTIQUITY
2. THEHARROW
The plough implies the use of an implement to break the clods
and render the land fit for seeding. The simplest and most primitive
form of harrow was doubtless the ‘ bush-harrow’ ; though in the
passage in Vergil,
Rastris glaebas qui frangit inertes,
Vimineasque trahit
there is some doubt as to the interpretation of ‘ crates ’, which some
consider means ‘ bush-harrow ’ (the ‘ arbuteae crates ’ of 1. 166), while
others take it to mean ‘ osier hurdles ’ 2 4 In any case, a bush-harrow
is a most ineffectual implement, as I have proved from using one ; and
even primitive man would soon devise something better. Martyn’s
interpretation of ‘ crates ’ as ‘ hurdles ’ suggests that the rectangular
implement with four cross-bars in Burkitt, plate 28, fig. I (fig. 6 unte)-
which he does not mention in the text-may be a heavy hurdle used as
a harrow.a6 On the other hand, it might conceivably (though perhaps
with less probability) be regarded as a very early form of drag-harrow,
with wooden teeth set in the beams, like the Roman irpex, ‘ Genus
rastrorum ferreorum, quod plures habent dentes ad extirpandas herbas
in agris ’.28 The shape is not mentioned ; such harrows are generally
triangular, and with iron or wooden teeth do quite good work. In the
earliest representation known to me of this type of harrow (14th
century), four transverse bars have teeth as well as the
OR TERRACES
3. LYNCHETS
The formation of lynchets, or as we now call them, terraces, is one
of the objects of a coffee-planter in cultivating his plantation (Fig. 10).
I grant that our terraces are only a fraction of the width of the Celtic
lynchets, but the method employed, and the result, are the same;
and as Dr Clay says (p. 57) they can be formed only by ploughing the
lower furrow, so that all the soil is turned down-hill, bringing back the

z3 Georg. I, 95.
24 So Martyn.
z5 Incidentally, one of the men in fig. 6 appears to be twisting an ox’s tail, a common
method in Africa of inducing a lazy ox to move.
z8 Festus, p. 339, S.V. Irpices. ‘ A kind of iron rake with many teeth for tearing out
weeds in the fields ’.
27 In a ‘ Shepherd’s Calendar ’, in the Brit. Mus. ; figured in Ditchfield, Old ViZZuge
Life, p. 137.
334
 ANCIENT AGRICULTURE

FIG. 5. CARVING OF PLOUGH, ' DOLMEN DES MARCHANDS'. (Le Rouzic)

FIG.9. CARVINGS OF PLOUGHS, MANE-ER-H'ROEK. (Le Rouzic)

FIG. 10. TERRACES, KENYA COFFEE PLANTATION

A-B, natural slope

335
 ANTIQUITY
plough idle instead of ploughing the upper furrow. Nowadays we use
a ' one-way ' or ' hillside ' plough, which has a reversible share held in
place by a catch ; at the end of each furrow the share is swung under
the plough into position for the next furrow.
Our ' lynchets ', on a slope such as shown in fig. 10,have a negative
lynchet z ft. high ; this can be produced in about two years by 3 to 4
ploughings a year, the level strip being disc-harrowed after each plough-
ing. On a rough estimate, one might put the formation of a Celtic
lynchet 10ft. high and zoo ft. wide at about 40 years, if it received two
ploughings a year ; if it was only ploughed once a year the time would
be longer. Once a coffee terrace has been properly formed, it does
not need to be ploughed with a ' one-way ' plough every time, as an
ordinary plough will preserve the proper level. So prehistoric man
probably did not find it necessary to plough the lower furrow only every
time, once his lynchet was established.
Our object in levelling the slope by terracing is to stop soil-wash,
the worst thing in nature with which the planter has to contend ; and
I can see no other reason for the ancient lynchets. The mere fact that
ground is sloping does not make it unsuitable for growing corn in wet
climates, as Dr Clay says (p. 58). For in East Africa, native fields of
maize, millet, and the short-stalked eleusine coracana, are often found
on very steep slopes ; and as long as they are prepared by hand (with
' hoes ') the soil washes very little in comparison with ploughed land.
It is when the plough comes into use that soil-wash starts ; hence the
necessity of flattening the slope. The formation of lynchets therefore
implies that soil-wash occurred, and that they were deliberately formed
to prevent it.28
4. MANURING AND ROTATION
It is not quite clear what Dr Curwen means by his statement that
there is ' no visible evidence of two-field or three-field rotation of
crops ' (p. 286), for a number of fields side by side need not imply any
sort of rotation. Even if the Britons marled their land with chalk?s
and the Welsh spread wood-ash on their fields,sO there are no grounds
for supposing that they had any sort of rotation ; for ' the Romans

28 D r Curwen considers that lynchets were not formed intentionally (p. 273).

29Pliny,N.H. XVII, 96, 4, seq.

30 Mabinogion ; quoted by Curwen, p. 287.

336
 ANCIENT AGRICULTURE
seem to have had some glimpses of rotation of crops ; but it does not
appear that any system of agriculture founded upon this knowledge
was in general use among them '.a1
Perhaps something can be learnt in this connexion from native
agricultural methods in Africa ; and the following account of Nand92
agriculture will perhaps suggest methods which may have been practised
by the more primitive of the early inhabitants of Britain. The Nandi
when preparing new ground for eleusine corn, break up the turf with
hoes. The turf is then collected into small heaps, and fired, and left to
smoulder till the grass is burnt The heaps of burnt earth are
then scattered over the ground, dug in with hoes, and a tilth is pre-
pared for the seed. The usual method of planting is to have a little of
everything in the same patch-eleusine, millet, maize, beans, and sweet
potatoes. Next year the stubble is dug over, and maize and millet
planted, but not eleusine ; for the Nandi consider that the ground must
be pared and burnt every year for this crop. So fresh ground is pre-
pared. After four or five years the patch is abandoned, and new ground
taken up. The cultivated fields are communal, and each man's holding
is a strip separated from the next holding by a narrow path.
We have here no rotation, but some idea of the manurial value of
wood-ash. The idea of leaving a patch of cultivated land after a few
years is due, not so much to a recognition of the benefit derived from
fallowing, as to the repressed nomadic instincts of the Nandi, for they
frequently move house as well. The natives in this part of the world
have not attained to the plough because a small area only need be
cultivated to support a family, and each man grows his own food. The
plough is the production of harder climates where a larger area is needed
under primitive methods of tillage than can conveniently be prepared
by hand.

31 Daubeny, Lectures on Roman Husbandry, p. 124,quoted by Conington, VergQ

(Bibl. Class.), I, 159.
32 I take the Nandi as an example although they are a pastoral tribe, because their
agricultural system (borrowed from the Bantu Kavirondo) is very well defined ; and
because I am much better acquainted with their methods than with those of other tribes.
33 The old-fashioned process of paring and burning formerly practised in England.

337