Here it is usual to enquire, whether it is necessary to attend to our numbers through the whole compass of a period, [Footnote: Our author here informs us, that what the Greeks called [Greek: periodos], a _period_, was distinguished among the Romans by the words _ambitus, circuitus, comprehensio, continuatio_, and _circumscriptio_. As I thought this remark would appear much better in the form of a note, than in the body of the work, I have introduced it accordingly.] or only at the beginning or end of it, or equally in both. In the next place, as _exact number_ seems to be one thing, and that which is merely _numerous_ another, it might be enquired wherein lies the difference. We might likewise consider whether the members of a sentence should all indifferently be of the same length, whatever be the numbers they are composed of;--or whether, on this account, they should not be sometimes longer, and sometimes shorter;--and when, and for what reasons, they should be made so, and of what numbers they should be composed;--whether of several sorts, or only of one; and whether of equal or unequal numbers;--and upon what occasions either the one or the other of these are to be used;-and what numbers accord best together, and in what order; or whether, in this respect, there is no difference between them;--and (which has still a more immediate reference to our subject) by what means our style may be rendered _numerous_. It will likewise be necessary to specify the rise and origin of a _periodical_ form of language, and what degree of compass should be allowed to it. After this, we may consider the members or divisions of a period, and enquire of how many kinds, and of what different lengths they are; and, if they vary in these respects, _where_ and _when_ each particular sort is to be employed: and, in the last place, the _use_ and application of the whole is to be fully explained;--a very extensive subject, and which is capable of being accommodated not only to one, but to many different occasions. But without adverting to particulars, we may discuss the subject at large in such a manner as to furnish a satisfactory answer in all subordinate cases.