This pronunciation is not universal over the Highlands. In some parts the _c_ retains its proper sound in all situations.

If the articulation in question had, from the first, been compounded, it is highly probable that it would have been represented, in writing, by a combination of letters, such as _chc_; especially as we find that the same sound is represented at other times, not by a single consonant, but by a combination, as in the case of _chd_. Why should it be thought that boc _a buck_, and bochd _poor_, were originally pronounced alike, when they are distinguished both in writing and signification?

The word [Hebrew: SHQ] _a sack_, has been transplanted from the Hebrew into many languages, among the rest the Gaelic, where it has been always written sac, although now pronounced sachc. In none of the other languages in which the word is used (except the Welsh alone), has the final palatal been aspirated. It would appear therefore that the sound sachc is a departure from the original Gaelic pronunciation. The same change may have happened in the pronunciation of other words, in which the plain _c_ is now aspirated, though it may not have been so originally.

[17] Though _th_ be quiescent in the middle of a polysyllable, over the North and Central Highlands, yet it is, with more propriety, pronounced, in the West Highlands, as an aspiration; as, athair _father_, mathanas _pardon_, pronounced a-hair, mahanas.

[18] I am informed that this pronunciation of _chd_ is not universal; but that in some districts, particularly the East Highlands, the _d_ has here, as in other places, its proper lingual sounds. In many, if not all the instances in which _chd_ occurs, the ancient Irish wrote _ct_. This spelling corresponds to that of some foreign words that have a manifest affinity to Gaelic words of the same signification; which, it is therefore presumable, were all originally pronounced, as they were written, without an aspiration, such as,

_Latin._ _Old French._ _Gaelic._

Noct-u Noct-is, &c. Nuict an nochd, _to night_.

Oct-o Huict Ochd, _eight_.

Benedict-um Benoict Beannachd, _blessing_.

Maledict-um Maudict Mallachd, _cursing_.

Ruct-us Bruchd, _evomition_.

Intellect-us Intleachd, _contrivance_.

Lact-is, -i, &c. Lachd, _milk_.

Dict-o, -are, &c. Deachd, _to dictate_.

Rego } Rect-um } Reachd, _a law, institution_.

From the propensity of the Gaelic to aspiration, the original _c_ was converted into _ch_, and the words were written with _cht_, as in the Irish acht _but_, &c., or with the slight change of _t_ into _d_, as in ochd, &c.

This is the opinion of O'Brien, when he says the word lecht is the Celtic root of the Latin _lectio_--the aspirate _h_ is but a late invention.--_O'Br. Ir. Dict. voc. lecht._ In process of time the true sound of _cht_ or _chd_ was confounded with the kindred sound of _chc_, which was commonly, though corruptly, given to final c.

[19] It is certain that the natural sound of d aspirated is that of [the Saxon ] or _th_ in _thou_; as the natural sound of _t_ aspirated is that of _th_ in _think_. This articulation, from whatever cause, has not been admitted into the Gaelic, either Scottish or Irish, although it is used in the kindred dialects of Cornwall and Wales.

[20] In sean _old_, the _n_ has its _plain_ sound when the following word begins with a Lingual. Accordingly it is often written in that situation seann; as, seann duine _an old man_, an t-seann tiomnaidh _of the old Testament_.

[21] So in Latin, _canmen_ from _cano_ was pronounced, and then written _carmen_; _genmen_ from the obsolete [Greek: geno] passed into _germen_.

[22] Another mode, proposed by a learned correspondent, of marking the distinction in the sound of the initial Linguals, is by writing the letter double, thus ll, nn, rr, when its sound is the same with that which is represented by those double letters in the end of a syllable; and when the sound is otherwise, to write the letter single; as, llamh _hand_, llion _fill_, mo lamh _my hand_, lion mi _I filled_.

It is perhaps too late, however, to urge now even so slight an alteration as this in the Orthography of the Gaelic, which ought rather to be held as fixed beyond the reach of innovation, by the happy diffusion of the Gaelic Scriptures over the Highlands.

[23] _Leathan re Leathan, is Caol re Caol._

Of the many writers who have recorded or taken notice of this rule, I have found none who have attempted to account for its introduction into the Gaelic. They only tell that such a correspondence between the vowels ought to be observed, and that it would be improper to write otherwise. Indeed, none of them seem to have attended to the different effects of a broad and of a small vowel on the sound of an adjacent consonant. From this circumstance, duly considered, I have endeavoured to derive a reason for the rule in question, the only probable one that has yet occurred to me.

[24] As deanuibh or deanaibh _do ye_, beannuich or beannaich _bless_.

[25] It is worthy of remark that in such words as caird-eil _friendly_, slaint-eil _salutary_, the substitution of _e_ in place of _a_ in the termination, both misrepresents the sound, and disguises the derivation of the syllable. The sound of this termination as in fear-ail _manly_, ban-ail _womanly_, is properly represented by _ail_. This syllable is an abbreviation of amhuil _like_, which is commonly written in its full form by the Irish, as fear-amhuil, &c. It corresponds exactly to the English termination _like_, in _soldier-like_, _officer-like_, which is abridged to _ly_, as _manly_, _friendly_. By writing _eil_ instead of _ail_, we almost lose sight of amhuil altogether.

[26] From the extracts of the oldest Irish manuscripts given by Lhuyd, Vallancey, and others, it appears that the rule concerning the correspondence of vowels in contiguous syllables, was by no means so generally observed once as it is now. It was gradually extended by the more modern Irish writers, from whom, it is probable, it has been incautiously adopted by the Scottish writers in its present and unwarrantable latitude.

The rule we have been considering has been reprobated in strong terms by some of the most judicious Irish philologers, particularly O'Brien, author of an Irish Dictionary printed at Paris 1768, and Vallancey, author of an Irish Grammar, and of various elaborate disquisitions concerning Irish antiquities, from whom I quote the following passages: "This Rule [of dividing one syllable into two by the insertion of an aspirated consonant]

together with that of substituting small or broad vowels in the latter syllables, to correspond with the vowel immediately following the consonant in the preceding syllable, has been very destructive to the original and radical purity of the Irish language." _Vallancey's Ir. Gram. Chap. III.

letter A._ "Another [Rule] devised in like manner by our bards and rhymers, I mean that which is called _Caol le caol, agus Leathan le leathan_, has been woefully destructive to the original and radical purity of the Irish language. This latter (much of a more modern invention than the former, for our old manuscripts show no regard to it) imports and prescribes that two vowels, thus forming, or contributing to form, two different syllables, should both be of the same denomination or class of either broad or small vowels, and this without any regard to the primitive elementary structure of the word." _O'Brien's Ir. Dict. Remarks on A._ "The words _biran_ and _biranach_ changed sometimes into _bioran_ and _bioranach_ by the abusive rule of _Leathan le leathan_." _Id. in voc._ Fear. The opinion of Lhuyd on this point, though not decisive, yet may properly be subjoined to those of Vallancey and O'Brien, as his words serve at least to show that this judicious philologer was no advocate for the Rule in question. "As for passing any censure on the rule concerning broad and small vowels, I chose rather to forbear making any remark at all upon them, by reason that old men who formerly wrote arget _silver_, instead of airgiod as we now write it, never used to change a vowel but in declining of words, &c. And I do not know that it was ever done in any other language, unless by some particular persons who, through mistake or ignorance, were guilty of it."

_Archaeol. Brit. Preface to Ir. Dict. translated in Bp. Nicolson's Irish Historical Library._

[27] Pinkerton's Inquiry into the History of Scotland.

[28] _E.g._, troidh _a foot_, has been written troidh or troigh, either of which corresponds to the pronunciation, as the last consonant is quiescent.

In Welsh, the articulation of the final consonant has been preserved, and the word is accordingly written troed. This authority seems sufficient to determine the proper orthography in Gaelic to be troidh and not troigh. For a like reason, perhaps, it would be proper to write traidh _shore_, rather than traigh, the common way of spelling the word, for we find the Irish formerly wrote traidh, and the Welsh traeth. Claidheamh _a sword_, since the final articulation was wholly dropped, has been sometimes written claidhe. The mode of writing it still with a final labial, though quiescent, will probably be thought the more proper of the two, when it is considered that claidheamh is the cognate, or rather the same word with the Irish cloidheamh the Welsh cleddyf, and the French glaive.

[29] I flatter myself that all my readers, who are acquainted with any of the ancient or the modern languages which have a distinction of gender in their attributives, will readily perceive that the import of the term Gender, in the grammar of those languages, is precisely what I have stated above. The same term has been introduced into the grammar of the English Tongue, rather improperly, because in an acceptation different from what it bears in the grammar of all other languages. In English there is no distinction of gender competent to Articles, Adjectives, or Participles.

When a noun is said to be of the masculine gender, the meaning can only be that the object denoted by it is of the male sex. Thus in the English grammars, gender signifies a quality of the _object_ named, while in other grammars it signifies a quality of the _name_ given to the object. The varieties of _who_, _which_, and _he_, _she_, _it_, refer not to what is properly called the _gender_ of the antecedent _noun_, but to the _Sex_ real or attributed, or the _absence of Sex_, of the _object_ signified by the antecedent. This is in effect acknowledged by writers on rhetoric, who affirm that in English the pronouns _who_, _he_, _she_, imply an express personification, or attribution of life, and consequently of Sex, to the objects to which these pronouns refer. The same thing is still more strikingly true of the variations on the termination of nouns, as _prince_, _princess_; _lion_, _lioness_, which are all discriminative of Sex. It seems therefore to be a mis-stated compliment which is usually paid to the English, when it is said that "this is the only language which has adapted the gender of its nouns to the constitution of Nature." The fact is, that it has adapted the _Form_ of some of the most common names of living creatures, and of a few of its pronouns, to the obvious distinction of _male_, and _female_, and _inanimate_, while it has left its nouns without any mark characteristic of _gender_. The same thing must necessarily happen to any language by abolishing the distinction of masculine and feminine in its attributives. If all languages had been constructed on this plan, it may confidently be affirmed that the grammatical term _gender_ would never have come into use. The compliment intended, and due to the English, might have been more correctly expressed, by saying that "it is the only language that has rejected the unphilosophical distinction of gender, by making its attributives, in this respect, all indeclinable."

[30] Uan beag bainionn, 2 Sam. xii. 3. Numb. vi. 14. So leomhann boirionn, Ezek. xix. 1.

[31] It must appear singularly strange that any nouns which signify females exclusively should be of the masculine gender. The noun bainionnach, is derived from the adjective bainionn, _female_, which is formed from bean, the appropriate term for a _woman_. Yet this noun bainionnach, or boirionnach, _a female_, is masculine, to all grammatical intents and purposes. We say boirionnach cir, _a civil woman_, am boirionnach maiseach, _the handsome woman_.

The gender of this Noun seems to have been fixed, not by its signification, but by its determination, for most Derivatives in _ach_ are masculines; as, oganach _a young man_, marcach _a horseman_, Albanach _a Scotsman_, &c. So in Latin, mancipium, scortum, though applied to persons, follow the gender of their termination.

[32] It was necessary to be thus explicit in stating the changes at the beginning and those on the termination as unconnected independent _accidents_, which ought to be viewed separately; because many who have happened to turn their thoughts toward the declension of the Gaelic noun have got a habit of conjoining these, and supposing that both contribute their united aid toward the forming the _cases_ of nouns. This is blending together things which are unconnected, and ought to be kept distinct. It has therefore appeared necessary to take a separate view of these two _accidents_ of nouns, and to limit the term _case_ to those changes which are made on the termination, excluding entirely those which take place at the beginning.

[33] It is to be observed that these names of the cases are adopted merely because they are already familiar, not because they all denominate correctly the relations expressed by the cases to which they are respectively applied. There is no Accusative or Objective case in Gaelic different from the Nominative; neither is there any Ablative different from the Dative. For this reason, it is not only unnecessary, but erroneous, to reckon up six Cases in Gaelic, distinguished not by the form of the Noun, but by the Prepositions prefixed. This is to depart altogether from the common and proper use of the term _Case_. And if the new use of that term is to be adopted, then the enumeration is still incomplete, for we ought to have as many Cases as there are Prepositions in the language. Thus, besides a Dative do Bhard, and an Ablative o Bhard, we should have an Impositive Case air Bhard, a Concomitative le Bard, an Insertive ann am Bard, a Precursive roimh Bhard, &c. &c. Grammarians have very correctly reckoned only five Cases in Greek, two in English, one in French [See _Moore_, _Murray_, _Buffier_, &c.] because the variations in the form of the Noun extend no further. Surely nothing but an early and inveterate prepossession in favour of the arrangements of Latin Grammar could ever have suggested the idea of Six Cases in Gaelic or in English.

[34] It is not improbable that anciently all feminine nouns, except a few irregular ones, added a syllable to the nominative, as _e_ or _a_, in forming the genitive. The translators of the S. S. have sometimes formed the genitive of feminine polysyllables in this manner, as sionagoige from sionagog, Mark v. 36, 38. But it appears more agreeable to the analogy of inflection that such polysyllables should now be written without an _e_ in the genitive.

[35] It is probable that this noun should rather be written adh. See McFarlane's Paraphrases, III. 3. also Lhuyd and O'Brien, _in loco_.

[36] Derivatives in _an_, and _ag_ should form their genitive according to the general Rule, _ain_, _aig_; and in pronunciation they do so. When the syllable preceding the termination ends in a small vowel, the Rule of 'Caol re caol' has introduced an _e_ into the final syllable, which is then written _ean_, _eag_. In this case writers have been puzzled how to form the genitive. The terminations _eain_, _eaig_, would evidently contain too many vowels for a short syllable. To reduce this awkward number of vowels they have commonly thrown out the _a_, the only letter which properly expressed the vocal sound of the syllable. Thus from caimean m. a _mote_, they formed the gen. sing. caimein; from cuilean m. a _whelp_, g. s.

cuilein; from duileag f. a _leaf_, g. s. duileig; from caileag f. a _girl_, g. s. caileig. Had they not yielded too far to the encroachments of the Rule of 'Caol re caol' they would have written both the nom. and the gen.

of these and similar nouns more simply and more justly, thus: caiman, g. s.

caimain; cuilan, g. s. cuilain; duilag, g. s. duilaig; cailag, g. s.

cailaig.

[37] In many instances, the Plural termination _a_ is oftener written with this final _n_ than without it. When the vowel preceding the termination is small, the termination _a_ or _an_ is very needlessly written _e_ or _ean_, to preserve the correspondence of vowels.

[38] We are informed by E. O'C. that this is the usual construction in the Irish Dialect, and it appears to be the same in the Scottish. Thus, air son mo dha shul, _for my two eyes_.--Judg. xvi. 28. Ir. & Scott. versions.

[39] So in Hebrew, we find a noun in the singular number joined with _twenty_, _thirty_, _a hundred_, _a thousand_, &c.

[40] The Pronouns tu _thou_, se _he_, si _she_, siad _they_, are not employed, like other nominatives, to denote the object after a transitive verb. Hence the incorrectness of the following expression in most editions of the Gaelic Psalms: Se chrunas _tu_ le coron graidh, Psal. ciii. 4., which translated literally signifies, _it is he whom thou wilt crown_, &c.

To express the true sense, viz., _it is he who will crown thee_, it ought to have been, se chrunas _thu_ le coron graidh. So is mise an Tighearn a slanuicheas _thu_, _I am the Lord that healeth thee_, Exod. xv. 26; Ma ta e ann a fhreagaireas _thu_, _If there be any that will answer thee_, Job v.

1; Co e a bhrathas thu? _Who is he that will betray thee?_ John xxi. 20., Comp. Gen. xii. 3. and xxvii. 29.

[41] This use of the Pronoun of the 2d person plural is probably a modern innovation, for there is nothing like it found in the more ancient Gaelic compositions, nor in the graver poetry even of the present age. As this idiom seems, however, to be employed in conversation with increasing frequency, it will probably lose by degrees its present import, and will come to be used as the common mode of addressing any individual; in the same manner as the corresponding Pronouns are used in English, and other European languages.

[42] There seems hardly a sufficient reason for changing the _d_ in this situation into _t_, as has been often done, as t'oglach for d'oglach _thy servant_, &c. The _d_ corresponds sufficiently to the pronunciation, and being the constituent consonant of the pronoun, it ought not to be changed for another.

[43] The Irish are not so much at a loss to avoid a _hiatus_, as they often use na for a _his_; which the translators of the Psalms have sometimes judiciously adopted; as,

An talamh tioram le na laimh Do chruthaich e 's do dhealbh. Psal. xcv. 5.

[44] In the North Highlands this Pronoun is pronounced sid.

[45] This Pronoun occurs in such expressions as an deigh na chuala tu _after what you have heard_; their leat na th' agad, or na bheil agad, _bring what you have_. It seems to be contracted for an ni a _the thing which_.

[46] There is reason to think that ge b'e is corruptly used for cia b' e.