OBS. 1.—­Grammarians in general treat of ellipsis without defining it; and exhibit such rules and examples as suppose our language to be a hundred-fold more elliptical than it really is.[479] This is a great error, and only paralleled by that of a certain writer elsewhere noticed, who denies the existence of all ellipsis whatever. (See Syntax, Obs. 24th on Rule 22d.) Some have defined this figure in a way that betrays a very inaccurate notion of what it is:  as, “ELLIPSIS is when one or more words are wanting to complete the sense.”—­Adam’s Lat. and Eng.  Gram., p. 235; Gould’s, 229.  “ELLIPSIS is the omission of one or more words necessary to complete the sense.”—­Bullions, Lat.  Gram., p. 265.  These definitions are decidedly worse than none; because, if they have any effect, they can only mislead.  They absurdly suggest that every elliptical sentence lacks a part of its own meaning!  Ellipsis is, in fact, the mere omission or absence of certain suggested words; or of words that may be spared from utterance, without defect in the sense.  There never can be an ellipsis of any thing which is either unnecessary to the construction or necessary to the sense; for to say what we mean and nothing more, never can constitute a deviation from the ordinary grammatical construction of words.  As a figure of Syntax, therefore, the ellipsis can only be of such words as are so evidently suggested to the reader, that the writer is as fully answerable for them as if he had written them.

OBS. 2.—­To suppose an ellipsis where there is none, or to overlook one where it really occurs, is to pervert or mutilate the text, in order to accommodate it to the parser’s or reader’s ignorance of the principles of syntax.  There never can be either a general uniformity or a self-consistency in our methods of parsing, or in our notions of grammar, till the true nature of an ellipsis is clearly ascertained; so that the writer shall distinguish it from a blundering omission that impairs the sense, and the reader or parser be barred from an arbitrary insertion of what would be cumbrous and useless.  By adopting loose and extravagant ideas of the nature of this figure, some pretenders to learning and philosophy have been led into the most whimsical and opposite notions concerning the grammatical construction of language.  Thus, with equal absurdity, Cardell and Sherman, in their Philosophic Grammars, attempt to confute the doctrines of their predecessors, by supposing ellipses at pleasure.  And while the former teaches, that prepositions do not govern the objective case, but that every verb is transitive, and governs at least two objects, expressed or understood, its own and that of a preposition:  the latter, with just as good an argument, contends that no verb is transitive, but that every objective case is governed by a preposition expressed or understood.  A world of nonsense for lack of a definition!