to duplication, than any others. These letters are double, not only at the end of those monosyllables which have but one vowel, as staff, mill, pass; but also under some other circumstances. According to general usage, final f is doubled after a single vowel, in almost all cases; as in bailiff, caitiff, plaintiff, midriff, sheriff, tariff, mastiff: yet not in calif, which is perhaps better written caliph. Final l, as may be seen by Rule 8th, admits not now of a duplication like this; but, by the exceptions to Rule 4th, it is frequently doubled when no other consonant would be; as in travelling, grovelling; unless, (contrary to the opinion of Lowth, Walker, and Webster,) we will have fillipping, gossipping, and worshipping, to be needful exceptions also.
Obs. 12.--Final s sometimes occurs single, as in alas, atlas, bias; and especially in Latin words, as virus, impetus; and when it is added to form plurals, as verse, verses: but this letter, too, is generally doubled at the end of primitive words of more than one syllable; as in carcass, compass, cuirass, harass, trespass, embarrass. On the contrary, the other consonants are seldom doubled, except when they come under Rule 3d. The letter p, however, is commonly doubled, in some words, even when it forms a needless exception to Rule 4th; as in the derivatives from fillip, gossip, and perhaps also worship. This letter, too, was very frequently doubled in Greek; whence we have, from the name of Philip of Macedon, the words Philippic and Philippize, which, if spelled according to our rule for such derivatives, would, like galloped and galloper, siruped and sirupy, have but one p. We find them so written in some late dictionaries. But if fillipped, gossipped, and worshipped, with the other derivatives from the same roots, are just and necessary exceptions to Rule 4th, (which I do not admit,) so are these; and for a much stronger reason, as the classical scholar will think. In our language, or in words purely English, the letters h, i, j, k, q, v, w, x, and y, are, properly speaking, never doubled. Yet, in the forming of compounds, it may possibly happen, that two Aitches, two Kays, or even two Double-ues or Wies, shall come together; as in withhold, brickkiln, slowwoorm, bayyarn.
OBS. 13.--There are some words--as those which come from metal, medal, coral, crystal, argil, axil, cavil, tranquil, pupil, papil--in which the classical scholar is apt to violate the analogy of English derivation, by doubling the letter l, because he remembers the ll of their foreign roots, or their foreign correspondents. But let him also remember, that, if a knowledge of etymology may be shown by spelling metallic, metalliferous, metallography, metallurgic, metallurgist, metallurgy, medallic, medallion, crystallize, crystalline, argillous, argillaceous, axillar, axillary, cavillous, cavillation, papillate, papillous, papillary, tranquillity, and pupillary, with double l, ignorance of it must needs be implied in spelling metaline, metalist, metaloid, metaloidal, medalist, coralaceous, coraline, coralite, coralinite, coraloid, coraloidal, crystalite, argilite, argilitic, tranquilize, and pupilage, in like manner. But we cannot well double the l in the former, and not in the latter words. Here is a choice of difficulties. Etymology must govern orthography. But what etymology? our own, or that which is foreign? If we say, both, they disagree; and the mere English scholar cannot know when, or how far, to be guided by the latter. If a Latin diminutive, as papilla from papula or papa, pupillus from pupus, or tranquillus from trans and quietus, happen to double an l, must we forever cling to the reduplication, and that, in spite of our own rules to the contrary? Why is it more objectionable to change pupillaris to pupilary, than pupillus to pupil? or, to change tranquillitas to tranquility, than tranquillus to tranquil? And since papilous, pupilage, and tranquilize are formed from the English words, and not directly from the Latin, why is it not as improper to write them with double l, as to write perilous, vassalage, and civilize, in the same manner?
OBS. 14.--If the practice of the learned would allow us to follow the English rule here, I should incline to the opinion, that all the words which I have mentioned above, ought to be written with single l. Ainsworth exhibits the Latin word for coral in four forms, and the Greek word in three. Two of the Latin and two of the Greek have the l single; the others double it. He also spells "coraliticus" with one l, and defines it "A sort of white marble, called coraline." [120] The Spaniards, from whose medalla, we have medal; whose argil[121] is arcilla, from the Latin argilla; and to whose cavilar, Webster traces cavil; in all their derivatives from these Latin roots, metallum, metal--coralium, corallium, curalium, or corallum, coral--crystallus or crystallum, crystal--pupillus, pupil--and tranquillus, tranquil--follow their own rules, and write mostly with single l: as, pupilero, a teacher; metalico, metalic; corolina (fem.) coraline; cristalino, crystaline; crystalizar, crystalize; traquilizar, tranquilize; and tranquilidad, tranquility. And if we follow not ours, when or how shall the English scholar ever know why we spell as we do? For example, what can he make of the orthography of the following words, which I copy from our best dictionaries: equip', eq'uipage; wor'ship, wor'shipper;--peril, perilous; cavil, cavillous;[122]--libel, libellous; quarrel, quarrelous;--opal, opaline; metal, metalline;[123]--coral, coralliform; crystal, crystalf
