Etymology Dictionary

BRIEF NOTES UPON THE LANGUAGES CITED

ENGLISH. Words marked (E.) are pure English, and form the true basis of the language. They can commonly be traced back for about a thousand years, but their true origin is altogether pre-historic and of great antiquity. Many of them, such as father, mother, &c., have corresponding cognate forms in Sanskrit, Greek, and Latin. These forms are collateral, and the true method of comparison is by placing them side by side. Thus father is no more 'derived' from the Sanskrit pitá¹ than the Skt. pitá is 'derived' from the English father. Both are descended from a common Aryan type, and that is all. Sometimes Sanskrit is said to be an 'elder sister' to English; the word 'elder' would be better omitted. Sanskrit has doubtless suffered less change, but even twin sisters are not always precisely alike, and, in the course of many years, one may come to look younger than the other. The symbol + is particularly used to call attention to collateral descent, as distinct from borrowing or derivation. English forms belonging to the 'Middle-English' period are marked 'M. E.' This period extends, roughly speaking, from about 1200 to 1460, both these dates being arbitrarily chosen. Middle-English consisted of three dialects, Northern, Midland, and Southern; the dialect depends upon the author cited. The spellings of the 'M.E.' words are usually given in the actual forms found in the editions referred to, not always in the theoretical forms as given by Stratmann, though these are, etymologically, more correct. Those who possess Stratmann's Dictionary will do well to consult it.

Words belonging to English of an earlier date than about 1150 or 1200 are marked 'A. S.', i.e. Anglo-Saxon. Some have asked why they have not been marked as 'O. E., i.e. Oldest English. Against this, there are two reasons. The first is, that 'O. E.' would be read as 'Old English,' and this term has been used so vaguely, and has so often been made to include 'M. E.' as well, that it has ceased to be distinctive, and has become comparatively useless. The second and more important reason is that, unfortunately, Oldest English and Anglo-Saxon are not coextensive. The former consisted, in all probability, of three main dialects, but the remains of two of these are very scanty. Of Old Northern, we have little left beyond the Northumbrian versions of the Gospels and the glosses in the Durham Ritual; of Old Midland, almost the only scrap preserved is in the Rushworth gloss to St. Matthew's Gospel; but of Old Southern, or, strictly, of the old dialect of Wessex, the remains are fairly abundant, and these are commonly called Anglo-Saxon. It is therefore proper to use 'A. S.' to denote this definite dialect, which, after all, represents only the speech of a particular portion of England. The term is well-established and may therefore be kept; else it is not a particularly happy one, since the Wessex dialect was distinct from the Northern or Anglian dialect, and 'Anglo-Saxon' must, for philological purposes, be taken to mean Old English in which Anglian is not necessarily included.

¹ Given as pitri in the Dictionary, this being the 'crude form' under which it appears in Benfey.

Anglo-Saxon cannot be properly understood without some knowledge of its phonology, and English etymology cannot be fairly made out without some notion of the gradations of the Anglo-Saxon vowel-system. For these things, the student must consult Sweet's Anglo-Saxon Reader and March's Grammar. Only a few brief hints can be given here.

SHORT VOWELS: a, æ, e, i o u, y.

LONG VOWELS: á, ǽ, é, í, ó, ú, ý.

DIPHTHONGS: eá, answering to Goth. au; eó, Goth. iu; also (in early MSS.) ie and ié.

BREAKINGS. The vowel a commonly becomes ea when preceded by g, c, or sc, or when followed by l, r, h, or x. Similarly e or i may become eo. The most usual vowel-change is that produced by the occurrence of i (which often disappeared) in the following syllable. This changes the vowels in row (1) below to the corresponding vowels in row (2) below.

These two rows should be learnt by heart, as a knowledge of them is required at almost every turn. Note that á and ǽ most often arise from an original (Aryan) i; whilst eó, eá, ú, and ý arise from original u.

Modern E. th is represented by A.S. þ or ð, used indifferently in the MSS.; see note to Th.

Strong verbs are of great importance, and originated many derivatives; these derivatives can be deduced from the form of the past tense singular, of the past tense plural, or of the past participle, as well as from the infinitive mood. It is therefore necessary to ascertain all these leading forms. Ex: bindan, to bind; pt. t. band, pl. bundon, pp. bunden. From the pt. t. we have the sb. band or bond; from the pp. we have the sb. bundle.

Examples of the Conjugations are these.

1. Feallan, to fall; pt. t. feóll, pl. feóllon; pp. feallen. Base FAL = ✔SPAR.

2. Bindan, to bind; pt. t. band, pl. bundon; pp. bunden. Base BAND = ✔BHANDH.

3. Beran, to bear; pt. t. bœr, pl. bóeron; pp. boren. Base BAR = ✔BHAR.

4. Gifan, to give; pt. t. geaf, pl. geáfon, pp. gifen. Base GAB.

5. Scínan, to shine; pt. t. scán, pl. scinon, pp. scinen. Base SKI.

6. Beódan, to bid; pt. t. beád, pl. budon, pp. boden. Base BUD.

7. Faran, to fare; pt. t. fór, pl. fóron, pp. faren. Base FAR = ✔PAR.

Strong verbs are often attended by secondary or causal verbs; other secondary verbs are formed from substantives. Many of these ended originally in -ian; the i of this suffix often disappears, causing gemination of the preceding consonant. Thus we have habban, to have (for haf-ian *); þeccan, to thatch (for þac-ian *); biddan, to pray (for bid-ian *); secgan, to say (for sag-ian *); sellan, to give, sell (for sal-ian *); dyppan, to dip (for dup-ian *); settan, to set (for sat-ian *). With a few exceptions, these are weak verbs, with pt. t. in -ode, and pp. in -od.

Authorities: Grein, Ettmuller, Somner, Lye, Bosworth, Leo, March, Sweet, Wright's Vocabularies.

OLD LOW GERMAN. Denoted by 'O. Low G.' This is a term which I have employed for want of a better. It is meant to include a not very large class of words, the precise origin of which is wrapped in some obscurity. If not precisely English, they come very near it. The chief difficulty about them is that the time of their introduction into English is uncertain. Either they belong to Old Friesian, and were introduced by the Friesians who came over to England with the Saxons, or to some form of Old Dutch or Old Saxon, and may have been introduced from Holland, possibly even in the fourteenth century, when it was not uncommon for Flemings to come here. Some of them may yet be found in Anglo-Saxon. I call them Old Low German because they clearly belong to some Old Low German dialect; and I put them in a class together in order to call attention to them, in the hope that their early history may receive further elucidation.

DUTCH. The introduction into English of Dutch words is somewhat important, yet seems to have received but little attention. I am convinced that the influence of Dutch upon English has been much underrated, and a closer attention to this question mighty throw some light even upon English history. I think I may take the credit of being the first to point this out with sufficient distinctness. History tells us that our relations with the Netherlands have often been rather close. We read of Flemish mercenary soldiers being employed by the Normans, and of Flemish settlements in Wales, 'where (says old Fabyan, I know not with what truth) they remayned a longe whyle, but after, they sprad all Englande ouer.' We may recall the alliance between Edward III and the free towns of Flanders; and the importation by Edward of Flemish weavers. The wool used by the cloth-workers of the Low Countries grew on the backs of English sheep; and other close relations between us and our nearly related neighbours grew out of the brewing-trade, the invention of printing, and the reformation of religion. Caxton spent thirty years in Flanders (where the first English book was printed), and translated the Low German version of Reynard the Fox. Tyndale settled at Antwerp to print his New Testament, and was strangled at Vilvorde. But there was a still closer contact in the time of Elizabeth. Very instructive is Gascoigne's poem on the Fruits of War, where he describes his experiences in Holland; and every one knows that Zutphen saw the death of the beloved Sir Philip Sidney. As to the introduction of cant words from Holland, see Beaumont and Fletcher's play entitled 'The Beggar's Bush.' After Antwerp had been captured by the Duke of Parma, 'a third of the merchants and manufacturers of the ruined city,' says Mr. Green, 'are said to have found a refuge on the banks of the Thames.' All this cannot but have affected our language, and it ought to be accepted, as tolerably certain, that during the fourteenth, fifteenth, and sixteenth centuries, particularly the last, several Dutch words were introduced into England; and it would be curious to enquire whether, during the same period, several English words did not, in like manner, find currency in the Netherlands. The words which I have collected, as being presumably Dutch, are deserving of special attention.

For the pronunciation of Dutch, see Sweet's Handbook of Phonetics. It is to be noted that the English oo in boor exactly represents the Dutch oe in boer (the same word). Also, that the Dutch sch is very different from the German sound, and is Englished by sc or sk, as in landscape, formerly landskip. The audacity with which English has turned the Dutch ui  in bruin (brown) into broo-in is an amazing instance of the influence of spelling upon speech. V and z are common, where English has f  and s. The symbol ij  is used for double i, and was formerly written y; it is pronounced like E. i  in wine. The standard Low German th appears as d; thus, whilst thatch is English, deck is Dutch. Ol appears as ou, as in oud, old, goud, gold, houden, to hold. D between two vowels sometimes disappears, as in weer (for weder *), a wether. The language abounds with frequentative verbs in -eren and -elen, and with diminutive substantives in -je (also -tje, -pje, -etje), a suffix which has been substituted for the obsolete diminutive suffix -ken.

Authorities: Oudemans, Kilian, Hexham, Sewel, Ten Kate, Delfortrie; dictionary printed by Taugchnitz.

OLD FRIESIC. Closely allied to Anglo-Saxon; some English words are rather Friesian than Saxon

Authorities: Richthofen; also (for modern North Friesic) Outzen; (for modern East Friesic) Koolman.

OLD SAXON. The old dialect of Westphalia, and closely allied to Old Dutch. Authority: Heyne.

LOW GERMAN. This name is given to an excellent vocabulary of a Low German dialect, in the work commonly known as the Bremken Wörterbuch.

SCANDINAVIAN. By this name I denote the old Danish, introduced into England by the Danes and Northmen who, in the early period of our history, came over to England in great numbers. Often driven back, they continually returned, and on many occasions made good their footing and remained here. Their language is best represented by Icelandic, owing to the curious fact that, ever since the first colonisation of Iceland by the Northmen about A.D. 874, the language of the settlers has been preserved with but slight changes. Hence, instead of its appearing strange that English words should be borrowed from Icelandic, it must be remembered that this name represents, for philological purposes, the language of those Northmen, who, settling in England, became ancestors of some of the very best men amongst us; and as they settled chiefly in Northumbria and East Anglia, parts of England not strictly represented by Anglo-Saxon, 'Icelandic' or 'Old Norse' (as it is also called) has come to be, it may almost be said, English of the English. In some cases, I derive 'Scandinavian' words from Swedish, Danish, or Norwegian; but no more is meant by this than that the Swedish, Danish, or Norwegian words are the best representatives of the 'Old Norse' that I could find. The number of words actually borrowed from what (in the modern sense) is strictly Swedish or strictly Danish is but small, and they have been duly noted.

Icelandic. Vowels, as in Anglo-Saxon, are both short and long, the long vowels being marked with an accent, as á, é, &c. To the usual vowels are added ö, and the diphthongs au, ey, ei; also æ, which is written both for æ and œ, strictly of different origin; also ja, já, jö, jó, jú. Among the consonants are ð, the voiced th (as in E. thou), and þ, the voiceless th (as in E. thin). D was at one time written both for d and ð. Þ, æ, and ö come at the end of the alphabet. There is no w. The A.S. w and hw  appear as v  and hv. The most usual vowel-change is that which is caused by the occurrence of i (expressed or understood) in the following syllable; this changes the vowels in row (1) below into the corresponding vowels in row (2) below.

Assimilation is common; thus dd  stands for ðd, or for Goth. zd (= A.S. rd ); kk, for nk; ll, for lr or lþ; nn, for nþ, nd, or nr; tt, for dt, ht, kt, nt, ndt, tþ. Initial sk should be particularly noticed, as most E. words beginning with sc or sk are of Scand. origin; the A.S. sc being represented by E. sh. Very remarkable is the loss of v  in initial vr = A.S. wr; the same loss occurring in modern English. Infinitives end in -a or -ja; verbs in -ja, with very few exceptions, are weak, with pp. ending in -ð, -ðr, -t, -tr, &c.; whereas strong verbs have the pp. in -inn.

Authorities: Cleasby and Vigfusson, Egilsson, Möbius, Vigfusson's Icelandic Reader.

Swedish. To the usual vowels add å, ä, ö, which are placed at the end of the alphabet. Diphthongs do not occur, except in foreign words. Qv is used where English has qu. The Old Swedish w (= A.S. w) is now v. The Icelandic and A.S. initial þ (= th) is replaced by t, as in Danish, not by d, as in Dutch; and our language bears some traces of this peculiarity, as, e.g. in the word hustings (for husthings), and again in the word tight or taut (Icel. þéttr).

Assimilation occurs in some words, as in finna (for finda *), to find, dricka (for drinka *), to drink; but it is less common than in Icelandic.

Infinitives end in -a; past participles of strong verbs in -en; weak verbs make the pt. t. in -ade, -de, or -te, and the pp. in -ad, -d, or -t.

Authorities: Ihre (Old Swedish, also called Suio-Gothic, with explanations in Latin); Widegren; Tauchitz dictionary; Rietz (Swedish dialects, a valuable book, written in Swedish).

Danish. To the usual vowels add æ and ö, which are placed at the end of the alphabet. The symbol ö is also written and printed as o with a slanting stroke drawn through it; thus ø. Qv  is used where English has qu; but is replaced by kv  in Aasen's Norwegian dictionary. V  is used where English has w. The Icelandic and A.S. initial þ (th) is replaced by t, as in Swedish; not by d, as in Dutch. Assimilation occurs in some words, as in drikke, to drink, but is still less common than in Swedish. Thus the Icel. finna, Swed. finna, to find, is finde in Danish. Mand  (for mann *), a man, is a remarkable form. We should particularly notice that final k, t, p, and f  sometimes become g, d, b, and v respectively; as in bog, a book, rag-e, to rake, tag-e, to take; ged, a goat, bid-e, to bite, græd-e, to weep (Lowland Scotch greet); reb, a rope, grib-e, to grip or gripe, knib-e, to nip; liv, life, kniv, knife, viv, wife. Infinitives end in -e; the past participles of strong verbs properly end in -en, but these old forms are not common, being replaced (as in Swedish) by later forms in -et or -t, throughout the active voice.

Authority: Ferrall and Repp's Dictionary.

Norwegian. Closely allied to Danish.

Authority: Aasen's Dictionary of Norwegian dialects (written in Danish).

GOTHIC. The Gothic alphabet, chiefly borrowed from Greek, has been variously transliterated into Roman characters. I have followed the system used in my Mœso-Gothic Dictionary, which I still venture to think the best. It is the same as that used by Massmann, except that I put w  for his v, kw for his kv, and hw  for his hv, thus turning all his v's into w's, as every true Englishman ought to do. Stamm has the same system as Massmann, with the addition of þ for th (needless), and q for kw, which is not pleasant to the eye; so that he writes qaþ for kwath (i.e. quoth). J corresponds to the E. y. One peculiarity of Gothic must be particularly noted. As the alphabet was partly imitated from Greek, its author used gg and gk (like Gk. γγ, γκ) to represent ng and nk; as in tuggo, tongue, drigkan, to drink. The Gothic vowel-system is particularly simple and clear, and deserving of special attention, as being the best standard with which to compare the vowel-systems of other Teutonic languages. The primary vowels are a, i, u, always short, and e, o, always long. The two latter are also written ê, ô, by German editors, but nothing is gained by it, and it may be observed that this marking of the letters is theoretical, as no accents appear in the MSS. The diphthongs are ai, au, ei, and iu; the two former being distinguished, theoretically, into ai and ái, au and áu. March arranges the comparative value of these vowels and diphthongs according to the following scheme,
  

Hence we may commonly expect the Gothic ai, ei, to arise from an original I, and the Gothic iu, au, to arise from an original U. The Gothic consonant-system also furnishes a convenient standard for other Teutonic dialects, especially for all Low-German. It agrees very closely with Anglo-Saxon and English. But note that A.S. gifan, to give, is Gothic giban (base GAB), and so in other instances. Also ear, hear, berry, are the same as Goth. auso, hausjan, basi, shewing that in such words the E. r  is due to original s.

Authorities: Gabelentz and Löbe, Diefenbach, Schulze, Massmann, Stamm, &c. (See the list of authorities in my own Mœso-Gothic Glossary, which I have used almost throughout, as it is generally sufficient for practical purposes)¹.

¹ Let me note here that, for the pronunciation of Gothic, the student should consult my edition of the Gospel of St. Mark in Gothic, Oxford, 1882; in which the errors occurring on p. 288 of my Gothic Glossary are corrected.

GERMAN. Properly called High-German, to distinguish it from the other Teutonic dialects, which belong to Low-German. This, of all Teutonic languages, is the furthest removed from English, and the one from which fewest words are directly borrowed, though there is a very general popular notion (due to the utter want of philological training so common amongst us) that the contrary is the case. A knowledge of German is often the sole idea by which an Englishman regulates his 'derivations' of Teutonic words; and he is better pleased if he can find the German equivalent of an English word than by any true account of the same word, however clearly expressed. Yet it is well established, by Grimm's law of sound-shiftings, that the German and English consonantal systems are very different. Owing to the replacement of the Old High German p by the Mod. G. b, and other changes, English and German now approach each other more nearly than Grimm's law suggests; but we may still observe the following very striking differences in the dental consonants. 

These changes are best remembered by help of the words day, tooth, foot, thorn, German tag, zahn, fuss, dorn; and the further comparison of these with the other Teutonic forms is not a little instructive.

The number of words in English that are borrowed directly from German is quite insignificant, and they are all of late introduction. It is more to the purpose to remember that there are, nevertheless, a considerable number of German words that were borrowed indirectly, viz. through the French. Examples of such words are brawn, dance, gay, guard, halbert, &c., many of which would hardly be at once suspected. It is precisely in accounting for these Frankish words that German is so useful to the English etymologist. The fact that we are highly indebted to German writers for their excellent philological work is very true, and one to be thankfully acknowledged; but that is quite another matter altogether.

Authorities: Wackernagel, Flügel, E. Müller. (I have generally found these sufficient, from the nature of the case; especially when supplemented by the words of Diez, Fick, Curtius, &c. But there is a good M.H.G. Dictionary by Lexer, another by Benecke, Müller, and Zarncke; and many more.)

FRENCH. The influence of French upon English is too well known to require comment. But the method of the derivation of French words from Latin or German is often very difficult, and requires the greatest care. There are numerous French words in quite common use, such as aise, ease, trancher, to cut, which have never yet been clearly solved; and the solution of many others is highly doubtful. Latin words often undergo the most curious transformations, as may be seen by consulting Brachet's Historical Grammar. What are called 'learned' words, such as mobile, which is merely a Latin word with a French ending, present no difficulty; but the 'popular' words in use since the first formation of the language, are distinguished by three peculiarities: (1) the continuance of the tonic accent, (2) the suppression of the short vowel, (3) the loss of the medial consonant. The last two peculiarities tend to disguise the origin, and require much attention. Thus, in the Latin bonitatem, the short vowel i, near the middle of the word, is suppressed; whence F. bonté, E. bounty. And again, in the Latin ligare, to bind, the medial consonant g, standing betwen two vowels, is lost, producing the F. lier, whence E. liable.

The result is a great tendency to compression, of which an extraordinary but well known example is the Low Latin ætaticum, reduced to edage by the suppression of the short vowel i, and again to aage by the loss of the medial consonant d; hence F. âge, E. age.

One other peculiarity is too important to be passed over. With rare exceptions, the substantives (as in all the Romance languages) are formed from the accusative case of the Latin, so that it is commonly a mere absurdity to cite the Latin nominative, when the form of the accusative is absolutely necessary to shew how the French word arose. On this account, the form of the accusative is usually given, as in the case of caution, from L. cautionem, and in numberless other instances.

French may be considered as being a wholly unoriginal language, founded on debased Latin; but it must at the same time be remembered that, as history teaches us, a certain part of the language is necessarily of Celtic origin, and another part is necessarily Frankish, that is, Old High German. It has also clearly borrowed words freely from Old Low German dialects, from Scandinavian (due to the Normans), and in later times, from Italian, Spanish, &c., and even from English and many entirely foreign languages.

Authorities: Cotgrave, Palsgrave, Littré, Scheler, Diez, Brachet, Burguy, Roquefort, Bartsch.

OTHER ROMANCE LANGUAGES. The other Romance languages, i.e. languages of Latin origin, are Italian, Spanish, Portuguese, Provençal, Romansch, and Wallachian. English contains words borrowed from the first four of these, but there is not much in them that needs special remark. The Italian and Spanish forms are often useful for comparison with (and consequent restoration of) the crushed and abbreviated Old French forms. Italian is remarkable for assimilation, as in ammirare (for admirare) to admire, ditto (for dicto), a saying, whence E. ditto. Spanish, on the other hand, dislikes assimilation, and carefully avoids double consonants; the only consonants that can be doubled are c, n, r, besides ll, which is sounded as E. l  followed by y consonant, and is not considered as a double letter. The Spanish ñ is sounded as E. n followed by y consonant, and occurs in dueña, Englished as duenna. Spanish is also remarkable as containing many Arabic (Moorish) words, some of which have found their way into English. The Italian infinitives commonly end in -are, -ere, -ire, with corresponding past participles in -ato, -uto, -ito. Spanish infinitives commonly end in -ar -er, -ir, with corresponding past participles in -ado, -ido-, -ido. In all the Romance languages, substantives are most commonly formed, as in French, from the Latin accusative.

CELTIC. Words of Celtic origin are marked '(C.)'. This is a particularly slippery subject to deal with, for want of definite information on its older forms in a conveniently accessible arrangement. That English has borrowed several words from Celtic cannot be doubted, but we must take care not to multiply the number of these unduly. Again, 'Celtic' is merely a general term, and in itself means nothing definite, just as 'Teutonic' and 'Romance' are general terms. To prove that a word is Celtic, we must first shew that the word is borrowed from one of the Celtic languages, as Irish, Gaelic, Welsh, Cornish, or Breton, or that it is of a form which, by the help of these languages, can be fairly presumed to have existed in the Celtic of an early period. The chief difficulty lies in the fact that Welsh, Irish, Cornish, and Gaelic have all borrowed English words at various periods, and Gaelic has certainly also borrowed some words from Scandinavian, as history tells us must have been the case. We gain, however, some assistance by comparing all the languages of this class together, and again, by comparing them with Latin, Greek, Sanskrit, &c., since the Celtic consonants often agree with these, and at the same time differ from Teutonic. Thus the word bard is probably Celtic, since it appears in Welsh, Irish, and Gaelic; and again, the word down (2), a fortified hill, is probably Celtic, because it may be compared with the A.S. tún, a Celtic d answering to A.S. t. On the other hand, the W. hofio, to hover, appears to be nothing but the common M.E. hoven, to hover, derived from the A.S. hof, a dwelling, which appears in E. hov-el. We must look forward to a time when Celtic philology shall be made much more sure and certain than it is at present; meanwhile, the Lectures on Welsh Philology by Professor Rhys give a clear and satisfactory account of the values of Irish and Welsh letters as compared with other Aryan languages.

Some Celtic words have come to us through French, for which assistance is commonly to be had from Breton. A few words in other Teutonic languages besides English are probably of Celtic origin.

RUSSIAN. This language belongs to the Slavonic branch of the Aryan languages, and, though the words borrowed from it are very few, it is frequently of assistance in comparative philology, as exhibiting a modern form of language allied to the Old Church Slavonic. My principal business here is to explain the system of transliteration which I have adopted, as it is one which I made out for my own convenience, with the object of avoiding the use of diacritical marks. The following is the Russian alphabet, with the Roman letters which I use to represent it. It is sufficient to give the small letters only.
  

This transliteration is not the best possible, but it will suffice to enable any one to verify the words cited in this work by comparing them with a Russian dictionary. I may here add that, in the 'Key' preceding the letter A, I have given Heym's dictionary as my authority, but have since found it more convenient to use Reiff (1876). It makes no difference. It is necessary to add one or two remarks.

The symbol ъ only occurs at the end of a word or syllable, and only when that word or syllable ends in a consonant; it is not sounded, but throws a greater stress upon the consonant, much as if it were doubled; I denote it therefore merely by an apostrophe. The symbol ъ most commonly occurs at the end of a word or syllable, and may be treated, in general, as a mute letter. э only occurs at the beginning of words, and is not very common. е may be represented by e at the beginning of a word, or otherwise by é, if necessary, since it cannot then be confused with з. It is to be particularly noted that j  is to have its French value, not the English; seeing that ж has just the sound of the French j, it may as well be so written. и and і are distinguished by the way in which they occur; ie can be written ié, to distinguish it from ie = ѣ. ѳ, which is rare, can be written as ph, to distinguish it from ф, or f; the sound is all one. By kh, Russ. х, I mean the German guttural ch, which comes very near to the sound of the letter; but the combinations ts, ch, sh, shch are all as in English. ы, or ui, resembles the French oui. The combinations ie, iu, ia, are to be read with i  as English y, i.e. yea, you, yaa. v, or y, pronounced as E. ee, is of no consequence, being very rare. I do not recommend the scheme for general use, but only give it as the one which I have used, being very easy in practice.

The Russian and Slavonic consonants agree with Sanskrit, Greek, and Latin rather than with Teutonic. The same may be said of Lithuanian, which is a very well preserved language, and often of great use in comparative philology. The infinitive mood of Russian regular verbs ends in -ate, -iate, iete, -ite, -uite, -ote, -ute; that of irregular verbs in -che, or -ti. In Lithuanian, the characteristic suffix of the infinitive is -ti.

SANSKRIT. In transliterating Sanskrit words, I follow the scheme given in Benfey's Dictionary, with slight modifications. The principal change made is that I print Roman letters instead of those which, in Benfey are printed with a dot beneath; thus I print ri, rí, t, th, d, dh, n, instead of ṛi, ṛí, ṭ, ṭh, ḍ, ḍh, ṇ. This is an easy simplification, and occasions no ambiguity. For ग, I print ç, as in Benfey, instead of ś, as in Monier Williams' Grammar. It might also be printed as a Roman s; but there is one great advantage about the symbol ç, viz. that it reminds the student that this sibilant is due to an original k, which is not slight advantage. The only letters that cause any difficulty are the four forms of n. Two of these, n and ṇ (or n), are easily provided for. ञ is represented in Benfey by ń, for which  I print ñ, as being easier; ङ is represented by ṅ, which I retain. The only trouble is that, in Monier Williams' Grammar, these appear as ṅ and n·, which causes a slight confusion.

Thus the complete alphabet is represented by a, á, i, í, u, ú, ri, rí, l ri, l rí, e, ai, o, au; gutturals, k, kh, g, gh, ṅ; palatals, ch, chh, j, hj, ñ; cerebrals, t, th, d, dh, n; dentals, t, th, d, dh, n; labials, p, ph, b, bh, m; semivowels, y, r, l, v; sibilants, ç, sh, s; aspirate, h. Add the nasal symbol m/tilde, and the final aspirate, h.

It is sometimes objected that the symbols ch, chh, are rather clumsy, especially when occurring as chchh; but as they are perfectly definite and cannot be mistaken, the mere appearance to the eye cannot much matter. Some write c and ch, and consequently cch instead of chchh; but what is gained in appearance is lost in distinctness; since च is certainly ch, whilst c gives the notion of E. c in can.

The highly scientific order in which the letters of the Sanskrit alphabet is arranged should be observed; it may be compared with the order of letters in the Aryan alphabet, given at p. 730, col. 2.

There are a few points about the values of the Sanskrit letters too important to be omitted. The following short notes will be found useful.

The Skt. ri answers to Aryan AR, and is perfectly distinct from r. Thus rich, to shine = Aryan ARK; but rich, to leave = Aryan RIK. An Aryan K becomes Skt. k, kh, ch, ç; Aryan G becomes g, j; Aryan GH becomes gh, h; Aryan T becomes t, th; Aryan P becomes p, ph; Aryan s becomes s and sh. See the table of 'Regular Substitution of Sounds' in Curius, i. 158. Other languages sometimes preserve a better form than Skt.; thus the ✔ AG, to drive, gives Lat. ag-ere, Gk. ἄγ-ειν, and (by regular change from g  to k ) Icelandic ak-a; but the Skt. is aj, a weakened form. The following scheme, abridged from Curtius, shews the most useful and common substitutions.
    

Both in this scheme, and at vol. i. p. 232, Curtius omits the Latin f  as the equivalent of Gk. χ initially. But I think it may fairly be inserted, since Gk. χολή = Lat. fel, Gk. χρίειν = Lat. friare, and Gk. χέειν is allied to Lat. fundere, on his own showing. Initial h is, however, more common, as in Lat. hiare, pre-hendere, humus, ansėr (for hanser *), hiems, heluus, haruspex, allied respectively to Gk. χαίνειν, χανδάνειν, χαμαί, χήν, χιών, χλόη, χολάδες. It becomes a question whether we ought not also to insert 'initial g ' in the same place, since we have Lat. grando and gratus, allied to Gk. χάλαζα and χαίρειν.

To the above list of substitutions may be added that of l  for r, which is a common phenomenon in nearly all Aryan languages; the comparison of Lat. grando with Gk. χάλαζα, has only just been mentioned. Conversely, we find r  for l, as in the well-known example of F. rossignol = Lat. lusciniola.

Authorities: Benfey; also (on comparative philology), Curtius, Fick, Vaniček; and see Peile's Greek and Latin Etymology, Max Müller's Lectures on the Science of Language; &c.

NON-ARYAN LANGUAGES: HEBREW. The Hebrew words in English are not very numerous, whilst at the same time they are tolerably well known, and the corresponding Hebrew words can, in general, be easily found. I have therefore contented myself with denoting the alphabet beth, gimel, daleth, &c. by b, g, d, h, v, z, kh, t, y, k, l, m, n, s, ', p, ts, q, r, sh or s, t. This gives the same symbol for samech and sin, but this difficulty is avoided by making a note of the few instances in which samech occurs; in other cases, sin is meant. So also with teth and tau; unless the contrary is said, tau is meant. This might have been avoided, had the words been more numerous, by the use of a Roman s and t for samech and teth, the rest of the word being in italics. I put kh for cheth, to denote that the sound is guttural, not E. ch. I denote ayin by the mark '. The other letters can be readily understood. The vowels are denoted by a, e, i, o u, á, é, í, ó, ú.

ARABIC. The Arabic alphabet is important, being also used for Persian, Turkish, Hindustani, and Malay. But as the letters are variously transliterated in various works, it seemed to be the simplest plan to use the spellings given in Richardson's Arabic and Persian Dictionary (with very slight modifications), or in Marsden's Malay Dictionary; and, in order to prevent any mistake, to give, in every instance, the number of the page in Richardson or Marsden, or the number of the column in Palmer's Persian Dictionary; so that, if in any instance, it is desired to verify the word cited, it can readily be done. Richardson's system is rather vague, as he uses t  to represent ت and ط (and also the occasional ة); also s to represent ث س and ص; also h for ح and ه; z  for ذ ز ض and ظ; k  for ق and ك; and he denotes ayin by the Arabic character. I have got rid of one ambiguity by using q (instead of k ) for ق; and for ayin I have put the mark ', as in Palmer's Persian Dictionary. In other cases, the reader can easily tell which t, s, h, or z  is meant, if it happens to be an initial  letter (when it is the most important), by observing the number of the page (or column) given in the reference to Richardson's or Palmer's Dictionary. Thus in Richardson's Dictionary, pp. 349-477 contain ت; pp. 960-981 contain ط; pp. 477-487 contain ث; pp. 795-868 contain س; pp. 924-948 contain ص; pp. 548-588 contain ح; pp. 1660-1700 contain ه; pp. 705-712 contain ذ; pp. 764-794 contain ز; pp. 949-960 contain ض; and pp. 981-984 contain ظ. In Palmer's Dictionary, the same letters are distinguished as t  (coll. 121-159); ṭ  (coll. 408-416); ṡ (coll. 160, 161); s (coll. 331-370); ṣ (coll. 396-405); ḥ (coll. 191-207); h (coll. 692-712); ż (coll. 283-287); z (coll. 314-330); ẓ (coll. 405-408); and z (z with a dot above and a dot below ) (coll. 416-418). Palmer gives the complete alphabet in the form a  [á, i, &c.] b, p, t, ṡ, j, ch, ḥ, kh, d, ż, r, z, zh, s, sh, ṣ, ẓ, ṭ, z, ' , gh, f, ḳ  [which I have written as q ], k, g, l, m, n, w, h, y. It deserves to be added that Turkish has an additional letter, sághír nún, which I denote by ñ, occurring in the word yeñi, which helps to form the E. word janisary. 

In words derived from Hindi, Hindustani, Chinese, &c., I give the page of the dictionary where the word may be found, or a reference to some authority.

From An Etymological Dictionary of the English Language, by the Rev. Walter W. Skeat, Litt.D. LL.D. Edin., M.A. Oxon. Elrington and Bosworth Professor of Anglo-Saxon in the University of Cambridge. Second Edition, 1883.