can be of much weight either way.  Yet it is important to know to what words the rule is, or is not, applicable.  In considering this vexatious question about the duplication of l, I was at first inclined to admit that, whenever final l has become single in English by dropping the second l of a foreign root, the word shall resume the ll in all derivatives formed from it by adding a termination beginning with a vowel; as, beryllus, beryl, berylline.  This would, of course, double the l in nearly all the derivatives from metal, medal, &c.  But what says Custom?  She constantly doubles the l in most of them; but wavers in respect to some, and in a few will have it single.  Hence the difficulty of drawing a line by which we may abide without censure. Pu’pillage and pu’pillary, with ll, are according to Walker’s Rhyming Dictionary; but Johnson spells them pu’pilage and pu’pilary, with single l; and Walker, in his Pronouncing Dictionary, has pupilage with one l, and pupillary with two.  Again:  both Johnson’s and the Pronouncing Dictionary, give us medallist and metallist with ll, and are sustained by Webster and others; but Walker, in his Rhyming Dictionary, writes them medalist and metalist, with single l, like dialist, formalist, cabalist, herbalist, and twenty other such words.  Further:  Webster doubles the l in all the derivatives of metal, medal, coral, axil, argil, and papil; but writes it single in all those of crystal, cavil, pupil, and tranquil—­except tranquillity.

OBS. 16.—­Dr. Webster also attempts, or pretends, to put in practice the hasty proposition of Walker, to spell with single l all derivatives from words ending in l not under the accent.  “No letter,” says Walker, “seems to be more frequently doubled improperly than l.  Why we should write libelling, levelling, revelling, and yet offering, suffering, reasoning, I am totally at a loss to determine; and, unless l can give a better plea than any other letter in the alphabet, for being doubled in this situation, I must, in the style of Lucian, in his trial of the letter T, declare for an expulsion.”—­Rhyming Dict., p. x.  This rash conception, being adopted by some men of still less caution, has wrought great mischief in our orthography.  With respect to words ending in el, it is a good and sufficient reason for doubling the l, that the e may otherwise be supposed servile and silent.  I have therefore made this termination a general exception to the rule against doubling.  Besides, a large number of these words, being derived from foreign words in which the l was doubled, have a second reason for the duplication, as strong as that which has often induced these same authors to double that letter, as