OBS. 19.—­A relative pronoun, being the representative of some antecedent word or phrase, derives from this relation its person, number, and gender, but not its case.  By taking an other relation of case, it helps to form an other clause; and, by retaining the essential meaning of its antecedent, serves to connect this clause to that in which the antecedent is found.  No relative, therefore, can ever be used in an independent simple sentence, or be made the subject of a subjunctive verb, or be put in apposition with any noun or pronoun; but, like other connectives, this pronoun belongs at the head of a clause in a compound sentence, and excludes conjunctions, except when two such clauses are to be joined together, as in the following example:  “I should be glad, at least, of an easy companion, who may tell me his thoughts, and to whom I may communicate mine.”—­Goldsmith’s Essays, p. 196.

OBS. 20.—­The two special rules commonly given by the grammarians, for the construction of relatives, are not only unnecessary,[382] but faulty.  I shall notice them only to show my reasons for discarding them.  With whom they originated, it is difficult to say.  Paul’s Accidence has them, and if Dean Colet, the supposed writer, did not take them from some earlier author, they must have been first taught by him, about the year 1510; and it is certain that they have been copied into almost every grammar published since.  The first one is faulty, because, “When there cometh no nominative case between the relative and the verb, the relative shall [not always] be the nominative case to the verb;” as may be seen by the following examples:  “Many are the works of human industry, which to begin and finish are [say is] hardly granted to the same man.”—­Dr. Johnson’s Adv. to Dict. “They aim at his removal; which there is reason to fear they will effect.”—­“Which to avoid, I cut them off.”—­Shak., Hen.  IV.  The second rule is faulty, because, “When there cometh a nominative case between the relative and the verb, the relative shall [not always] be such case as the verb will have after it;” as may be seen by the following examples:  “The author has not advanced any instances, which he does not think are pertinent.”—­Murray’s Gram., i, 192. “Which we have reason to think was the case with the Greek and Latin.”—­Ib., 112.  “Is this your son, who ye say was born blind?”—­John, ix, 19.  The case of the relative cannot be accurately determined by any rules of mere location.  It may be nominative to a verb afar off, or it may be objective with a verb immediately following; as, “Which I do not find that there ever was.”—­Knight, on the Greek Alphabet, p. 31.  “And our chief reason for believing which is that our ancestors did so before us.”—­Philological Museum, i, 641.  Both these particular rules are useless, because the general rules for the cases, as given in chapter third above, are applicable to relatives, sufficient to all the purpose, and not liable to any exceptions.