preffed in words soft and flowing; and plain language devoid of ornament, is adapted to fubjects grave and didactic. Language may be confidered as the dress of thought; and where the one is not fuited to the other, we are sensible of incongruity, in the fame manner as where a judge is dreffed like a fop, or a peafant like a man of quality. The intimate connection that words have with their meaning, requires that both be in the fame tone. Or, to exprefs the thing more plainly, the impreffion made by the words ought as nearly as poffible to resemble the impreffion made by the thought. The fimilar emotions mix fweetly in the mind, and augment the pleasure *. On the other hand, where the impreffions made by the thought and the words are diffimilar, they are forc'd into a fort of unnatural union, which is disagreeable +. In the preceding chapter, concerning the language of paffion, I had occafion to give many examples of deviations from this * Chap. 2. part 4. +Ibid. rule Kk 2 1 rule with regard to the manner of expreffing passions and their fentiments. But as the rule concerns the manner of expreffing thoughts and ideas of all kinds, it has an extenfive influence in directing us to the choice of proper materials. In that view it must be branched out into several particulars. And I muft obferve, in the first place, that to write with elegance, it is not fufficient to exprefs barely the conjunction or disjunction of the members of the thought. It is a beauty to find a fimilar conjunction or disjunction in the words. This may be illuftrated by a familiar example. When we have occafion to mention the intimate connection that the foul has with the body, the expreffion ought to be the foul and body; because the particle the, relative to both, makes a connection in the expreffion, which refembles in fome degree the connection in the thought. But when the foul is diftinguished from the body, it is better to say the foul and the body, because the disjunction in the words resembles the disjunction in the thought. In the follow ing examples the connection in the thought is happily imitated in the expreffion. Conftituit agmen; et expedire tela animofque, equitibus juffis, &c. Again: Livy, 1. 38. § 25. Quum ex paucis quotidie aliqui eorum caderent aut vulnerarentur, et qui fuperarent, fefsi et corporibus et animis effent, &c. Livy, 1. 38. § 29. Poft acer Mneftheus adducto conftitit arcu, The following paffage of Tacitus appears to me not so happy. It approaches to wit by connecting in the foregoing manner things but flightly related, which is not altogether fuitable to the dignity or gravity of history. Germania omnis a Galliis, Rhætiifque, et Pannoniis, Rheno et Danubio fluminibus; a Sarmatis Dacifque, mutuo metu aut montibus feparatur. De moribus Germanorum. I I am more doubtful about this other instance: The fiend look'd up, and knew His mounted fcale aloft; nor more, but fled Murm'ring, and with him fled the fhades of night. Paradife Loft, B. 4. at the end. I shall add fome other examples where the oppofition in the thought is imitated in the words; an imitation that is diftinguished by the name of antithefis. Speaking of Coriolanus foliciting the people to be made conful: With a proud heart he wore his humble weeds. Coriolanus. Had you rather Cæfar were living, and die all flaves; than that Cæfar were dead, to live all free men ? Julius Cæfar. He hath cool'd my friends and heated mine ene mies. Shakespear. Why, if two gods fhould play fome heav'nly match, Pawn'd Pawn'd with the other; for the poor rude world: Merchant of Venice, at 3. fc. 6. This rule may be extended to be extended to govern the conftruction of fentences or periods. A fentence or period in language ought to exprefs one entire thought or mental propofition; and different thoughts ought to be separated in the expreffion by placing them in different fentences or periods. It is therefore offending againft neatness, to crowd into one period entire thoughts which require more than one; for this is conjoining in language things that are separated in reality; and confequently rejecting that uniformity which ought to be preferved betwixt thought and expreffion. Of errors against this rule take the following examples. Cæfar, defcribing the Suevi: Atque in eam fe confuetudinem adduxerunt, ut locis frigidiffimis, neque veftitus, præter pelles, habeant quidquam, quarum, propter exiguitatem, magna |