compact fentence, had they not been followed by the last member; but as that only adds to the fenfe of the preceding members, and does not qualify them, the whole affemblage of members, taken together, form but one loofe fentence.

The laft member of the laft fentence is neceffarily connected with what precedes, because it modifies or reftrains the meaning of it; for it is not meant, that the pleafures of the imagination do not fuffer the mind to fink into negligence and remiffnefs in general, but into that particular negligence and remiffnefs which is apt to accompany our more fenfual delights. The firft member of this fentence affords an opportunity of explaining this by its oppofite: for here it is not meant, that those pleasures of the imagination only are of this innocent nature which do not require fuch a bent of thought as is neceffary to our more ferious employments, but that of this nature are the pleasures of the imagination in general; and it is by afking the question whether a preceding member affirms any thing in general, or only affirms fomething as limited or qualified by what follows, that we fhall difcover whether thefe members are either immediately or remotely connected, and, confequently, whether they form a loofe or a compact fentence: as the. former member, therefore, of the last sentence, is not neceffarily connected with those that fucceed, the fentence may be pronounced to be a loofe fentence.

If these obfervations have any folidity, we have at last arrived at the true diftinction between a period and a loose sentence; which is,

that a period is an affemblage of fuch words, or members, as do not form fenfe independent on each other; or if they do, the former modify the latter, or inverfely; and that a loofe fentence is an afSemblage of fuch words or members as do form fenfe, independent on those that follow, and at the fame time are not modified by them: A period or compact fentence, therefore, is divifible into two kinds; the firft, where the former words and members depend for fenfe on the latter, as in the fentence, As we cannot difcern the shadow moving along the dial-plate, fo the advances we make in learning are only perceived by the distance gone over. Which, for diftinction's fake, we may call a direct period. The fecond kind of period, or compalt fentence, is that where, though the first part forms fenfe without the latter, it is nevertheless modified by it; as in the fentence, There are feveral arts which all men are in fome measure mafters of, without being at the pains of learning them. Which we may call an inverted period. The loofe fentence has its first members forming fenfe, without being modified by the latter; as in the fentence, Perfons of good taste expect to be pleafed at the fame time they are informed; and think that the best fenfe always deferves the best language. In which example, we find the latter member adding fomething to the former, but not modifying or altering it,

It will readily occur to the critical reader, that, in this definition of a period, I have departed widely from the doctrine of the ancients, who confider it as an affemblage of members, and not of words only; but as fuch a reader will know the difficulty of giving a precife idea of a period, according to the opinion of the

ancients, and what diverfity and uncertainty there is about it among the moderns; he will the more eafily excufe my hazarding a definition. of my own. My principal object has been, to give fuch a definition as would be clear, precife, and useful; fuch a one as would beft answer the purposes of pronunciation, by exactly drawing the line between the connection and disjunction of words, without making ufe of fuch indefinite terms as the more or lefs intimate connection of the parts, or the concurrence of the parts to the plenitude of a total fenfe.

Sentences thus defined and diftinguished into their feveral kinds, we fhall be better enabled to give fuch rules for dividing them by paufes, as will reduce punctuation to fome rational and fteady principles. Previously, however, to thefe rules, it will be neceffary to obferve, that as the times of the paufes are exceedingly indefinite, the fewer diftinctions we make between them, the lefs we fhall embarrass the reader : the common eftimate of the times of the comma, the femicolon, the colon, and the period, in the geometrical proportions of 1, 2, 4, 8, pleases us, from its analogy with the times of the femibref, minim, crotchet, and quaver in mufic; but every one will confefs, at first fight, that as these diftinctions in reading are arbitrary, they are useless; every one feels a difference between a greater and a smaller pause, but few can conceive degrees of these : I shall beg leave, therefore, to reduce the number of pauses to three; namely, the fmaller paufe, answering to the comma the greater paufe, answering to the femicolon and colon; and the greateft paufe, answering to the period. The ancients knew

D

nothing of the femicolon; and if we confider practice and real utility, I believe it will be found, that the three distinctions of the ancients anfwer every useful purpose in writing and reading.

The smaller pause, the greater pause, and the greatest pause, are the diftinctions, therefore, I shall beg leave to adopt in the rules to be given for dividing a sentence: and as the divifion of a fentence depends neceffarily on its ftructure, and the greater or lefs connection of its parts, it will be proper to begin with the direct period; that is, where no fenfe is formed till the fentence is concluded.

Rule I. Every direct period consists of two principal conftructive parts, between which parts the greater pause must be inferted; when thefe parts commence with conjunctions that correfpond with each other, they are fufficiently distinguishable; as in the following fentence:

As we cannot difcern the fhadow moving along the dial-plate, fo the advances we make in knowledge are only perceived by the distance gone over.

Here we may obferve, that the first conftructive part begins with as, and the fecond with fo the expectation is excited by the first, and anfwered by the latter at that point, therefore, where the expectation begins to be answered, and the fenfe begins to form, the principal pause is to be used; and, by these means, the two contrafted and correfpondent parts are diftinctly viewed by the mind.

A period may be direct, and its parts as neceffarily connected, where only the first conjunction is expreffed.

EXAMPLE.

As in my fpeculations I have endeavoured to extinguish paffion and prejudice, I am still defirous of doing fome good in this particular. Spectator.

Here the word fo is understood before I am, and the long paufe as much required as if fo had been expreffed; fince it is here the fentence naturally divides into two correfpondent and dependent parts.

That point, therefore, where the expectation begins to be answered, or where one part of the fentence begins to modify the other, is the point which we must be the most careful to mark; as it is here the fentence naturally divides into its principal conftructive parts.

Rule II. Every inverted period confifts of two principal conftructive parts, between which parts the greater pause must be inserted; these parts divide at that point, where the latter part of the sentence begins to modify the former; in periods of this kind, the latter conjunction only is expreffed, as in the example: Every one that fpeaks and reafons is a grammarian, and a logician, though he may be utterly unacquainted with the rules of grammar, or logic, as they are delivered in books and fyftems. If we invert this period, we' fhall find it fufceptible of the two correfpondent conjunctions though and yet; as, Though utterly unacquainted with the rules of grammar or logic, as delivered in books and fyftems, yet every man who speaks and reafons is a grammarian and a logician.-This inverfion of the order of a sentence, is, perhaps, the best criterion of the connection of its parts; and proves that the former, though forming complete fenfe by itself, is modified by the latter. Thus in