Latin, "little conclusion," a clausula is a poetic and rhetorical device used by ancient Greek and Roman orators like Cicero and Plato, for several practical and stylistic purposes. Romans did not think of sentences and paragraphs in the way we think of them now, as self-contained syntactic formations clearly separated from one another through the use of punctuation marks, and separated in speech by modulating one's vocal pitch (such as the rising inflection used at the ends of questions, to show one is asking a question). Instead, a sententia was an entire idea being addressed, even at such lengths as modern anglophones would call an "essay" rather than a "sentence," and the smaller divisions of this idea were the membra of that idea: its appendages, which formed its whole body. At the end of each membrum, a skilled orator (or one who wished to appear skilled, by emulating the stylings of those who actually were) would append a final remark or statement, the clausula, which would take on a specific poetic rhythm that served as a fingerprint or signature unique to that specific orator. The clausula rhythm would be firmly and loudly enunciated, compared to the other contents of the oration, allowing listeners at great distance to hear that rhythm and - if they were familiar with the orator's rhetoric - to recognise precisely who was giving a speech. It was not uncommon for a speech to be documented, in part or in whole, through memorisation by someone in attendance, who might then go on to relay it to other audiences from street corners, and the clausula's loud emphasis helped to draw an interested crowd.
By the time Cicero was an active participant in Roman politics, clausulae had proliferated in rhetoric, such that a clausula would follow every membrum of a speech, and not only the final membrum, resulting in the clausula's signature rhythm being hammered over and over, between short stretches of softer-spoken and less rhythmically structured speech. What differentiates a clausula from any other ending on a membrum is its repetition; if there is no significant recurrence of the same rhythm at the end of many membra, then they are not clausulae, but are only ordinary speech.
Clausulae were also used on written letters and messages transmitted through memorisation and recitation by a personal servant. In this way, a clausula acted as a means to sign a letter without actually placing one's name on the document itself, conferring plausible deniability to its author and recipient, if the letter should fall into the hands of an unsuspected foe.
Historians and classicists today derive great utility out of the clausulae of the ancients, because they allow fairly reliable identification of unsigned letters and untitled orations. It was typical in Cicero's time and later for an orator to change their signature clausula every few years, and this means that historians can use that change to pinpoint when an undated letter or speech was written, even when its contents are too ambiguous to plainly refer to any specific events with an already-known year of occurrence; this is especially useful for dating Plato's works. We can also use clausulae to track the shifting of political sympathies: some orators would use clausula which emulated another orator for a time, and later, following a discontinuity of their allegiance or common goals, that mimicry would not only stop, but the clausulae would change to be as dissimilar as possible from that of their former ally. The ancient orators of Rome (and Greece before her) had much to say on the topic of clausulae, further assisting modern historians in judging their intentions: Cicero's dē Ōrātōre and Aristotle's Ars Rhētorica address the topic extensively. The Roman rhetoric instructor Quintilian even warned his students not to make their clausulae too similar to the poetic meters used in verse, lest it become too tempting and too easy for bawdy and satirical poets to make popular songs in mockery of their oration. Accordingly, we must understand clausulae not just as a mark of skill and an identifying gesture, but - because they could associate an author to allies or foes, drawing ire or ridicule - a diplomatic ploy.
To a linguist, what is perhaps the most useful and fascinating clausula feature of all, is that they prove definitively how specific words were pronounced by specific orators. There are many Latin words with spellings that lend to the assumption that their primary syllable stress is placed on the second or third syllable, but we can discern from clausulae that - at least for a specific time, or by a specific author - their stress was actually on the first syllable, or could at least "get away with" being placed on the first syllable, without sounding uncouth. In instances where these "peculiar" pronunciations occur among multiple Romans whose careers were contemporary to one another, but whose social classes and hometowns were not the same as one another, we can reasonably suppose that the "peculiar" pronunciation was actually just how that word was properly spoken at the time, and was not any one orator's personal idiosyncrasy or regional accent bleeding through into their rhetoric. To the classics linguist hoping to to uncover more pronunciation quirks that haven't been documented in the usual educational style guides, like Benjamin Hall Kennedy's Revised Latin Primer, clausulae offer both utility and fun.
By the by, you might espy: this whole time I've used clausulae.
Count and see: in iambs three, paragraphs end similarly.
Haec pro certāmine "Nōdātōris Ferreōrum," annō MMXXIII, trīcēsima de trigintā sententiīs, scrīpta est.