or partial, to the rule for plurals in these cases is explained by reference to the paradigm of the Anglo-Saxon vowel declension. There was a class of nouns in the neuter and another in the feminine (the neuter more numerous than the feminine), that added nothing to the stem in either number to form the nominative and the accusative. Our unchanged plurals can be traced back directly to AngloSaxon, or can be charged to Anglo-Saxon analogy.

3. Plurals formed by Internal Change.-In English to-day six nouns, man, foot, tooth, goose, mouse, and louse, add neither s nor n to form the plural, nor is their plural like the singular. It is formed by a change of the stemvowel. This irregularity also is inherited from the AngloSaxon. The Anglo-Saxon fōt (foot) may be taken to represent them. (Anglo-Saxon ō and è are pronounced like long o and long a.)

This change of o to e in three cases was caused by the vowel i, which once followed the stem, and had sufficient influence upon the preceding syllable to modify its vowel in the direction of i. This variation is called mutation, and shows its striæ in many other English words. The i of the ending disappeared, as the change of vowel it had wrought in the stem was regarded as ample to mark the case. So that what was originally euphonic, accompanying the ending, and accidentally helping to denote the case, came in time to do it exclusively. But the cases constantly occurring were the nominative and the accusative (objective).

The dative singular fet faded from memory, and fet (the ē becoming, as usual, ee in English), whenever found, was looked upon as the plural of fot. The same i wrought the change which appears in the English plural of the remaining five, men, teeth, geese, mice, and lice. In the AngloSaxon there were eleven of these mutation plurals.

4. Foreign Plurals in English.-There are foreign nouns in English which have brought along their original endings. They end in us, like focus and fungus—plurals, foci and fungi; in um, as memorandum and stratum,plurals, memoranda and strata; in is,―oasis and parenthesis pluralizing as oases and parentheses; in ix, ex,—calix and vortex pluralizing variously as calices and calixes, vortices and vortexes. Hebrew cherub and seraph have the plurals cherubim and seraphim ; and French beau, the plural beaux. They all in time bow to the law which imposes s or es as the plural ending in English; but in the process of Anglicising they have two forms; as, seraphim and seraphs, indices and indexes. Frequently, as in these two pairs, these different forms have different meanings assigned them.

XVIII. The Loss of Grammatical Gender.-By grammatical gender we mean the gender of the noun as determined by its termination or declension, without exclusive, if indeed any, reference to the sex, or to the absence of sex, of the object named. We mean gender as it is in German and French to-day. In these, one must remember the gender of every noun, since its gender determines the form of the adjective used. From this labor the student of English is exempt, as the sex of the object determines the gender of its name. If the object is a male, the noun is masculine ; if a female, the noun is feminine; if without sex, neuter.

As with us neither of the articles nor any adjective changes its form on account of gender, a knowledge even of sex is essential only in determining which of the third personal pronouns he, she, or it, in the singular-should be used in place of the noun, if any is employed.

The total abandonment of grammatical gender by the Anglo-Saxon and the Norman-French nouns took place largely while the languages were uniting to form the English. The loss is an enormous gain in relieving the memory and in aiding personification.

CHAPTER V.

GRAMMATICAL CHANGES OF THE ANGLO-SAXON IN BECOM

ING ENGLISH.-THE ADJECTIVE.

XIX. Forms of the Adjective.-1. Its Declension.-In Anglo-Saxon, the adjective was inflected much as were the nouns of the vowel declension. But if a demonstrative or a possessive pronoun preceded the adjective, the terminations were precisely those of the noun in the n, or consonant, declension.

Of this cumbersome mass of endings, matched or a little overmatched by the modern German, traces may be seen in the e ending of Chaucer's adjectives in the plural. But by 1550, even this vowel, to which, as in the noun, the other vowels had weakened, disappeared, and the adjective became flectionless.

2. Its Comparison in er and est.—In our Indo-European family of languages, the comparative is formed by adding a syllable to the simple stem of the adjective; the superlative, by adding a suffix to the comparative. In the Teutonic member of the family, is or os was the suffix added to form the comparative; to this, ta was attached to form the superlative.

The s of the comparative is or os, except in worse and less, passed over into r. Professor Hadley in his scholarly essay On Passive Formations has paralleled this change of s to r, showing that the r in the passive of the Latin verb

-in laudor, laudaris, laudatur, laudamur, laudantur, for instance-is the s of the reflexive pronoun se (self) changed to r. To this essay the reader is referred for the additional support which an examination of other languages, in our family and out of it, gives to this assigned origin of the r of our comparatives.

The i or o of the comparative suffix of adjectives, but not of adverbs, was dropped in Anglo-Saxon; not, however, without occasionally having an effect upon the vowel of the preceding syllable, an effect still seen in elder and eldest, from the positive old. The s of the full superlative suffix ista or osta did not undergo rhotacism, but remained s. When now the i or the o of the Teutonic suffix, dropped in Anglo-Saxon, was resumed in early English; when the weakening of it to e, and of the i and o of the superlative to e, took place; and when the final a of that degree, softened to e, was dropped; then the adjective formed its comparative in er and the superlative in est, as now.

3. Its Comparison by Adverbs.—It is said that the first comparison by means of adverbs is found in the Ancren Riwle, about 1220—the meste dredful. This method is not Anglo-Saxon but Norman-French; though it is worth noting that the adverbs used in the comparison, more and most, less and least, are themselves compared in the old way. The adverbial comparison, used mainly with polysyllabic adjectives, and with participles employed as adjectives, has gained so rapidly upon the other that Trench predicted the extinction of the comparison in er and est. We do not, as they did in Shakespeare's and in Milton's day, write preposterousest, flourishingest, and dangerouser; but, when we find in writers like Hawthorne, Whately, Professor Whitney, Matthew Arnold, Lowell, Thackeray, Hadley, George Eliot,