The risks of errors in the transcription must certainly have been materially increased if in the larger publishing establishments the practice was followed of writing from dictation, one “reader” supplying simultaneous “copy” to a number of scribes. It seems probable that in no other way would it have been practicable to produce with sufficient speed and economy the editions required, and I find myself in accord with Birt in the conclusion that dictating was the method generally followed, at least in the more important establishments and for the larger editions. The scribes must of necessity have had a scholarly training, and ought also to have possessed some familiarity with the texts to which they were listening; while with the most skilful and scholarly scribes a careful revision of their copies would have been essential. Haenny is of opinion that dictation was rarely if ever employed. He lays stress on the fact that the term employed by Cicero in referring to the multiplication of copies was _describere_, and he contends that this stands simply for copying and cannot be translated as writing from dictation.[210] One indication of the size of the editions prepared of new books is given in the many references to the various uses found for the “remainders” or unsold copies. The most frequent fate of unsuccessful poetry was for the wrapping of fish and groceries, while large supplies of surplus stock found their way from the booksellers to the fires of the public baths.[211] Cooks also were large buyers of remainders of editions. An author who was voluminous and who had not been able to secure a publisher, might even, as the wags suggested, find it convenient to be burned upon a pile of his own manuscripts. It is evident that in these earlier days of publishing it was no easier than at present for authors or publishers to calculate with accuracy the extent of the public interest in their productions, while it is also probable that then as now an author would rather pay for the making of an abundant supply than incur the dreadful risk of not having enough copies to meet the immediate demand. While the Augustan age witnessed a decided development in the literary interests of the Roman community, and while the organization of such bookselling establishments as those of Atticus, Tryphon, and the Sosii gave to authors the needed machinery for bringing their writings before the public, it is probable that for the larger number of the writers of the time the receipts from the books were very inconsiderable.