There are two things, however, which give a relish to our language,--well- chosen words, and harmonious _numbers_. Words may be considered as the _materials_ of language, and it is the business of _number_ to smooth and polish them. But as in other cases, what was invented to serve our necessities was always prior to that which was invented for pleasure; so, in the present, a rude and simple style which was merely adapted to express our thoughts, was discovered many centuries before the invention of _numbers_, which are designed to please the ear. Accordingly _Herodotus_, and both his and the preceding age had not the least idea of prosaic _number_, nor produced any thing of the kind, unless at random, and by mere accident:--and even the ancient masters of rhetoric (I mean those of the earliest date) have not so much as mentioned it, though they have left us a multitude of precepts upon the conduct and management of our style. For what is easiest, and most necessary to be known, is, for that reason, always first discovered. Metaphors, therefore, and new-made and compounded words, were easily invented, because they were borrowed from custom and conversation: but _number_ was not selected from our domestic treasures, nor had the least intimacy or connection with common language; and, of consequence, not being noticed and understood till every other improvement had been made, it gave the finishing grace, and the last touches to the style of Eloquence. As it may be remarked that one sort of language is interrupted by frequent breaks and intermissions, while another is flowing and diffusive; it is evident that the difference cannot result from the natural sounds of different letters, but from the various combinations of long and short syllables, with which our language, being differently blended and intermingled, will be either dull and motionless, or lively and fluent; so that every circumstance of this nature must be regulated by _number_. For by the assistance of _numbers_, the _period_, which I have so often mentioned before, pursues it's course with greater strength and freedom till it comes to a natural pause. It is therefore plain that the style of an Orator should be measured and harmonized by _numbers_, though entirely free from verse; but whether these numbers should be the same as those of the poets, or of a different species, is the next thing to be considered. In my opinion there can be no sort of numbers but those of the poets; because they have already specified all their different kinds with the utmost precision; for every number may be comprized in the three following varieties:--_viz_. a _foot_ (which is the measure we apply to numbers) must be so divided, that one part of it will be either equal to the other, or twice as long, or equal to three halves of it. Thus, in a _dactyl_ (breve-macron-macron) (long-short-short) the first syllable, which is the former part of the foot, is equal to the two others, in the _iambic_