It is a short essay with references to the Nights by Henry James Pye, from 1792. (http://en.wikipedia.org/wiki/Henry_James_Pye)
Its title is "Note to Aristotle's Poetics" and I found it at this website on English poetry:
"EVEN IN THE EPOPEE THE FABLES SHOULD HAVE A DRAMATIC FORM, AND RELATE TO ONE ENTIRE AND COMPLEAT ACTION, WHICH HAS A BEGINNING, A MIDDLE, AND AN END.
It has been already observed that in every fictitious tale, independently of technical rules, it is impossible to keep up the attention and interest of the piece without confining the time of the fable within a certain boundary. This rule of nature is confirmed by the practice of all our good novel writers. I will not only instance the novels of Fielding, who as being a scholar, and rather fond of shewing he was, may be supposed to display his acquaintance with the precepts of the Stagirite, or rather the models from whence they are drawn: but those of Richardson, of Mrs. Smith, and of Miss Burney, who cannot be supposed to be influenced by any pedantry of this kind. Even in those novels which are written on what Dr. Beattie calls the historical plan, such as Peregrine Pickle and Roderick Random, though they begin with the infancy of the hero, they by no means compleat his life. The first events are rather preparatory to, than part of the main object of the story, the body of which, or to speak dramatically, the plot, is the love of the principal characters, as the solution of it, or catastrophe, is their marriage.
The mode used by the orientals to give a species of unity to their complicated fables is very singular. Of this the Arabian Nights exhibit a curious specimen. A general story, or ground-work for the whole, is first formed. This is the bloody vow of the sultan in consequence of the sultana's infidelity, the generous resolution of the vizir's daughter, and her final triumph. Into this the other stories are woven, but the introductory tale is continually brought to our recollection by the conversation that precedes the narrative of every night. Every story is besides branched out into a number of others, to each of which it serves as a common bond of union, as the original one is to the whole.
By this strange contrivance an appearance of general unity is kept up without the least of that effect which is proposed as the consequence of unity; as the mind is disagreeably perplexed by the broken chain of the narrative, expectation is suspended till all interest in the fable is lost, and instead of perspicuity confusion is produced. This arrangement is preserved in the first half of Mr. Galland's translation. In the latter part he has given such separate stories as struck him, without dividing the nights, or preserving any connexion between them, except the catastrophe of the leading fable. Mr. Andrews in his Anecdotes, gives a humorous reason for M. Galland's change of conduct. But I believe the principal cause of it was the length of the original work, which he has greatly abridged, as will be apparent on comparing the number of nights in that part of his translation where they are noticed, which is a full half of his work, with the compleat number of a thousand and one.
Dr. Beattie seems to question the authenticity of this work. I think the reason he urges; (the French features given it by M. Galland,) can have no weight with a person who has ever read a French translation from any language. Whoever will compare this work even through the medium of a French translation with the many WESTERN oriental tales, to which it has given birth, will see strong marks of original and real character. But I believe the authenticity of this work is capable of stronger proof. I have been informed that Professor White has a compleat copy in Arabic. Mr. Richardson also, in his Arabic Grammar, has printed one of the fables at length in the original language, and quoted verses from another which are not translated by M. Galland, though the tale from which they are taken is.
This mode of narration was adopted by Ariosto, and was copied from him by our countryman Spenser.
As for Ariosto, his imagination is so brilliant, his subject so wonderfully varied
From grave to gay, from lively to severe:
there is such a mingled vein of sublimity, and humour, running through the whole work, that notwithstanding the many absurdities it contains, and such a total want of connexion in the incidents, that to enable the reader at all to follow the thread of the scattered tales, the commentators have been obliged to have recourse to inartificial assistance or marginal references; yet we can hardly wish it to have been in any respect different from what it is. But as the work of Spenser is entirely of a serious cast, our taste is more fastidious, and indeed the attempt at uniformity, which is avowed by the author, causes us to be more disgusted both with the want of it in the Cantos that are preserved, and the apparent inadequacy of the whole plan proposed, had it been compleatly carried into execution. The unity produced by the introduction of a general kind of secondary hero pervading the whole, must have been very awkward and very uninteresting. Prince Arthur engaged as an assistant to the several allegorical heroes in their respective adventures, would have exactly resembled the pentathlete, as described by Plato, who, however skilful he might be in the contest with those who like himself were trained to the practice of various exercises, was always inferior to those athletes who applied themselves to one only, in that particular exercise."