| 1. | France. | The Gard. Joins the Rhone. |
| Germany. | Gard(aha), 8th cent. The Gart(ach). | |
| The Kart(haue) in Prussia. | ||
| 2. | With the ending en. | |
| Scotland. | The Gairden. Joins the Dee. | |
| France. | The Gardon. Joins the Rhone. | |
| Greece. | Jardanus ant. in Crete—here? | |
In the Gael. sgia, Welsh ysgw, guard, protection, and in the Welsh ysgi, separation or division, we have two senses, of which the latter may be more suitable for the following. The Editor of Smith's Ancient Geography suggests that the Scius of Herodotus may be the present Isker in Bulgaria: in an etymological point of view this seems probable, for as Scius = Welsh ysgi, so Isker = Welsh ysgar of the same meaning.
| Netherlands. | The Schie by Schiedam. |
| Danub. Prov. | Scius ant., now the Isker? |
From the Gael. scar, sgar, Welsh ysgar, Ang.-Sax. scêran, to divide, in the sense of boundary, may be the following. The small river Scarr in Dumfriesshire forms for six miles a boundary between different parishes.[65]
| 1. | England. | The Shere. Kent. |
| Scotland. | The Scarr. Dumfriesshire. | |
| The Shira. Argyle. | ||
| Germany. | Scere, 11th cent. The Scheer. | |
| 2. | With the ending en. | |
| England. | The Skerne. Durham. | |
| Germany. | Schyrne, 11th cent., not identified. | |
Any names in which the sense of land, terra, occurs, may, I think, be explained most reasonably in the sense of boundary or territorial division. To this Grimm places the Fulda of Germany, Fuld(aha), 8th cent., referring it to Old High Germ. fulta, Ang.-Sax. folde, earth, ground.
Perhaps also to a similar origin may be referred the Mold(au) in Bohemia, and the Mold(ava) of Moldavia. But the Gael. and Ir. malda, malta, gentle, slow, Anglo-Sax. milde, Eng. mild, may be perhaps more suitable: the Mulde, which joins the Elbe, and which in the 8th cent. appears as Milda, seems more probably from this origin.
The Bord(au), formerly Bordine, which forms for some distance the boundary between East and West Friesland, may, as suggested by Förstemann, be derived from Old Fries. and Anglo-Saxon bord, border. Another river of the same name (p. 33) may perhaps be otherwise derived.
I am inclined to bring in here the Granta, and to suggest that it may have been a Sax. or Angle name of the Cam, or of a certain part of the Cam. This river seems to have formed one of the boundaries of the country of the Gyrvii;[66] its name appears in Henry of Huntingdon as Grenta; and the Old Norse grend, Mod. Germ. grenze, boundary, seems a probable etymon.
[65] Statistical account of Scotland.
[66] See an article by the Rev. W. Stubbs on "The Foundation and early Fasti of Peterborough," in the Archæological Journal for Sept., 1861.
In this chapter I include some names which do not come under any of the foregoing heads, or which have been omitted in their places.
The following have generally been referred to Gael. caol, straight, narrow.
| 1. | England. | The Cole. Warwickshire. |
| The Coly. Devon. | ||
| 2. | With the ending en. | |
| England. | The Colne. Three rivers. | |
But even if this derivation is to be received, we must seek another meaning for the Kola in Russian Lapland, and the Koli(ma) in Siberia—the latter in particular being a large river, with a wide estuary.
The Gael. and Ir. beag, little, forms the ending of some Irish river-names, as the Awbeg, the Owenbeg, and the Arobeg.[67] The meaning in all these cases is "little river"—owen being the same as avon, aw the simple form av of the same word, and aro an appellative as at p. 38, now lost in the Celtic.
From the Gael. suail, small, have also been derived the Swale and other following rivers. Chalmers rightly objects to this as inconsistent with the character of the rivers, though the derivation which he proposes to substitute, from ys-wall, a sheltered place, affords, it must be admitted, no very happy alternative. I think the word contained must be related to Old High German swal, Old Norse svelgr, gurges, Eng. swell, though it is wanting in the Celtic.
| 1. | England. | The Swale. Two rivers, Kent and Yorkshire. |
| The Swily. Gloucestershire. | ||
| Ireland. | The Swelly. Donegal. | |
| The Swilly. Ulster. | ||
| Germany. | Suala ant. The Schwale. | |
| France. | Sulgas ant., now the Sorgue. | |
| Russia. | The Sula—here? | |
| 2. | With the ending en. | |
| Ireland. | The Sullane. | |
The following must be referred to Old High Germ. sualm, gurges, an extension of the previous word sual.
| Germany. | Sualman(aha), 8th century. The Schwalm. |
| Sulmana, 8th cent. The Sulm. | |
| Belgium. | The Salm. Prov. Liège. |
| France. | The Solman. Dep. Jura. |
The Shannon has by some writers been derived from Ir. sean or shean, old. But inasmuch as there is no river that is otherwise than old, the term could only be used in a poetic sense, like "that ancient river, the river Kishon." A more suitable etymon, however, seems to me to be found in Ir. and Obs. Gael. siona, delay; this corresponds with the Gaelic form of the name, Sionan, given by Armstrong.
| Scotland. | The Shin. Sutherland. |
| Ireland. | Senus (Ptolemy). The Shannon. |
| Germany. | Sinna, 8th cent. The Sinn. |
| Belgium. | The Senne. Joins the Dyle. |
| Italy. | Sena ant., now the Nevola. |
| Aust. Pol. | The San, two rivers—here? |
| India. | The Seena—here? |
From the Gael. cobhair, Ir. cubhair, foam, froth, appear to be the following.
| England. | The Cober. Cornwall. |
| The Cover. Yorkshire. | |
| Russia. | The Choper. |
| Asia. | Chaboras ant., now the Khabur—here? |
| India. | Chaberis ant., now the Caveri—here? |
From the Ir. and Obs. Gael. breath, pure, clear, I take to be the following.
| England. | The Bratha. Lake District. |
| Scotland. | The Broth(ock). Forfar. |
| Germany. | The Brett(ach). Joins the Kocher. |
| The Brat(awa) in Bohemia. | |
| Braht(aha),[68] 10th century. The Bracht—here? | |
| Asia Minor. | Practius ant.—here? |
And from the Ir. brag, running water, I follow Mone in taking the following.
| 1. | England. | The Bray. Devon. |
| Ireland. | The Bray. Wicklow. | |
| France. | The Bray. Joins the Loire. | |
| Germany. | The Brege, in the Scharwarzwald. | |
| 2. | With the ending en. | |
| England. | The Braine. Joins the Blackwater. | |
| Ireland. | Breagna, an old name for the Boyne. | |
A root for river-names, to which might be put the following, is found by Förstemann in Old High Germ. rôr, Mod. Germ. rohr, arundo, Eng. rush.
| Germany. | Ror(aha), 11th century, now the Rohrbach. |
| Rura, 8th cent. The Ruhr. | |
| Holland. | The Roer. Joins the Maas. |
The word sil in river-names would seem to have the meaning of still or sluggish water. The Gael. has sil, to drop, rain, drip; and the Arm. has sila, to filter. (The Old Fries. sil, canal, seems hardly a related word; it appears more probably to be connected with Old Norse sîla, to cut, to furrow.) According to Pliny, the Scythian name of the Tanais or Don was Silis; and several other Scythian rivers had the same name, (Grimm, Gesch. d. Deutsch. Sprach.) In this point of view the above derivation might seem too restricted, and we might think of sil, as of sal, (p. 75), as formed by the prefix s from the root al or il, to go, (p. 71), in the simple meaning of water. According to Strabo and Pliny the Silaris of Italy had the property of petrifying any plant thrown into it; but as, according to Cluvier, the modern inhabitants of its banks know nothing of any such property, it would rather seem as if the story had been made to fit the supposed connection of the name with silex, flint.
| 1. | Switzerland. | Sil(aha), 11th cent. The Sihl. |
| Italy. | Silis ant., now the Sile. | |
| Scotland. | The Shiel in Argyleshire—here? | |
| Germany. | The Schyl (ant. Tiarantus)—here? | |
| 2. | With the ending en. | |
| Sweden. | Siljan. Lake. | |
| Russia. | The Shelon—here? | |
| 3. | With the ending er. | |
| Naples. | Silaris ant., now the Silaro. | |
The form silv I take to be an extension of sil, similar to others previously noticed.
| 1. | Russia. | The Silva. Gov. Perm. |
| 2. | With the ending er. | |
| England. | The Silver. Devon. | |
The Simois in the Plain of Troy I have suggestively placed at p. 119 to Gael. saimh, slow, tranquil. But, taking the epithet lubricus applied to it by Horace, we might perhaps seek a stronger sense from the same root, as found in Welsh seimio, to grease, saim, tallow.
The water of the Liparis in Cilicia, according to Polyclitus, as quoted by Pliny, was of such an unctuous quality that it was used in place of oil. Probably only for the purpose of anointing the person, to which extent the story is confirmed by Vitruvius. Hence no doubt its name, from Sansc. lip, to be greasy, Gr. λιπαρος, unctuous.
Grimm (Gesch. d. Deutsch. Sprach.) suggests a similar origin for the Ister, p. 117, referring it to Old Norse istra, Dan. ister, fat, grease, Gr. στέαρ. He puts it, however, in a metaphorical sense, as "the fattening, fructifying river." With deference, however, to so high an authority, this explanation seems to me rather doubtful. For the ending ster, as I have elsewhere observed, is common to many river-names, and I have taken it to be, like the Arm. ster, formed by a phonetic t, from the Sansc. sri, to flow.
Also, from the root of the Sansc. sri, to flow, I take to be Gael. sruam, and again taking the phonetic t, the word stream, strom, common to all the Teutonic dialects. In these two forms we find the ancient names of two rivers—the Syrmus of Thrace, and the Strymon or Strumon, the present Struma, of Macedonia.
[67] The derivation at p. 120 I must retract, finding beg as a termination of other Irish river-names.
[68] Wiegand, (Oberhessische ortsnamen), refers this name to Old High Germ. braht, fremitus.
The names of rivers form a striking commentary on the history of language, so admirably expounded to the general reader in the recent work of Professor Max Müller.
When we review the long list of words that must have once had the meaning of water or river, we can hardly fail to be struck with the number that have succumbed in what he so aptly terms "the struggle for life which is carried on among synonymous words as much as among plants and animals."
We see too how large a portion of this long list of appellatives may ultimately be traced back to a few primary roots. And how even these few primary roots may perhaps be resolved into a still smaller number of yet more simple forms.
I take for instance, as a primitive starting point in river-names, the Sansc. root î, â, or ay, signifying to move, to flow, to go. We have appellatives even in this simple form, as the Old Norse â, Anglo-Sax. aê, water, river. But whether they directly represent the root, or whether, like the French eau, p. 30, they have only withered down to it again, after a process of germinating and sprouting, I do not take upon me to determine.
Then we have the roots, also of the kind called primary, ab, ar, ir, ag, ikh, il, it, all having the same general meaning, to move, to go, and from which, as elsewhere noticed, are also derived a number of appellatives for water or river in the various Indo-European languages. I should be inclined to suggest that the whole of these are formed upon, and are modifications of the simple root î, â, or ay, and that the following remarks made by Max Müller respecting secondary roots, may be extended also to them. "We can frequently observe that one of the consonants, in the Aryan languages, generally the final, is liable to modification. The root retains its general meaning, which is slightly modified and determined by the changes of the final consonants." He instances the Sansc. tud, tup, tubh, tuj, tur, tuh, tus, all having the same general meaning, to strike.
Again—there are forms such as ang, amb, and, &c., which are merely a strengthening of the roots ag, ab, ad, or at, and which also are found in a number of appellative forms.
We might pursue the subject still further, and enquire whether the secondary forms, such as sar, sal, car, cal, all having the same general meaning, to move, to go, may not be formed, by the prefix of a consonant, on the roots ar and al, and so also be ultimately referred to the simple root î or â.
As also the silent and ceaseless flow of water is the most natural and the most common emblem of the efflux of time; so in the same root is to be found the origin of many of the words which mean time and eternity. The Gr. αει, the Goth. aiv, the Anglo-Sax. awa, Eng. ever and aye, are all from this same root, so widely spread in river-names, and express the same idea which speaks—
P. 25.
To the root ab or ap, water, place the Lith. and Lett. uppe, river, whence the following.
| Germany. | The Oppa in Silesia. |
| Russia. | The Upa. Joins the Oka. |
| The Ufa. Joins the Bielaya. |
P. 33.
To the root ud place as an appellative the Obs. Gael. ad, water. And add to form No. 1 the following names.
| Russia. | The Uda. Gov. Kharkov. |
| France. | The Odde. Dep. Allier. |
P. 35.
The Celt. word and or ant, water, is nothing more than a strengthening of the above Obs. Gael. ad.
P. 40.
In referring to the root ark, erk, I have omitted the Ir. earc, water, the appellative most nearly concerned. The Basque erreca, brook, might be taken to be borrowed from the Celtic, did we not find in the same language the more primitive words ur and errio, p. 38, which seem to form a link with the Indo-European languages.
P. 49.
To the root nig, ni, place—
| 1. | France. | The Né. Joins the Charente. |
| Norway. | The Nia. Stift Trondjem. | |
| 3. | With the ending es. | |
| Russia. | The Nerussa. Gov. Orel. | |
P. 63.
To the root wig, wic, wy, place the two following names. The Welsh gwy, water, is the word most nearly concerned in most of the group.
| England. | The Wyck. Buckinghamshire. |
| Russia. | The Ui. Gov. Orenburg. |
P. 64.
To the root vip place as an appellative the Welsh gwibio, to rove, wander, gwibiau, serpentine course. Probably upon the whole the sense of tortuousness is that which should be recognized. The following name probably belongs to form No. 1.
| Spain. | The Quipar. Joins the Segura. |
P. 70.
The Celtic languages have a trace of the word trag, to run, in the Old Ir. traig, foot (Zeuss, Gramm. Celt.)
P. 83.
For
| Greece. | Pydaras ant. Thrace. |
Read
| Thrace. | Pydaras ant. |
P. 84.
To the Ir. biol, buol, water, place the following names.
| England. | The Beaulieu, also called the Exe, in Hampshire. |
| Scotland. | The Beauly. Inverness. |
| Italy. | Paulo ant., now the Paglione. |
P. 85.
I apprehend that in the opinion of Celtic scholars of the present day the Ancient British deity Cocidis is not considered to have any connection with the river Coquet.
P. 91.
It seems probable that the word asp in river-names is formed by metathesis from the word aps, p. 27, form 5.
P. 97.
The Gryffe and the Girvan may perhaps be better derived from the Gael. grib, swift.
P. 132.
To the root pad or pand, to spread, may probably be placed—
| England. | The Pant. Essex. |
P. 135.
From the root tan, may be derived the Dniester, (=Danaster), from ster, river. Or it might be from the root dan, as in Danube, p. 116.
P. 136.
The Dan. tang, sea-weed, does not seem to be connected with any word signifying water: it represents the Old Norse tag, twig.
P. 145.
To the root vind, white, clear, place—
| England. | The Wente. Yorkshire. |
P. 149.
To the Sansc. taras, Welsh têr, pure, clear, place—
| Thrace. | Tearus ant. |
(Ancient Names in Italics.)