Some time in 1668, after the book had been out for half a year or more, it seemed to Simmons to need a push; and Milton agreed, and helped him.
Yet one other fact about this famous First Edition. Not only are copies of it to be found with at least nine varieties of title-page, bearing date 1667, 1668, or 1669; and not only do these copies have or lack the fourteen pages of interpolated prose-matter (consisting of "The Argument," the Apology for the Verse, and the List of Errata), according as they were bound and issued after or before Simmons's effort in 1668 to "push" the book by that insertion; but minute differences may be found here and there in the printing of the poem itself in these various copies. Such differences are very rare indeed, and hardly ever of the least particle of consequence. By way of specimen, I may mention that, while a copy now before me of the date 1669, with Helder's name in the imprint, exhibits several miscountings of the lines in tens in the margin between line 810 and line 1010 of Book IX., no such errors are to be found in the corresponding margins of the present facsimile, taken from a copy of the first issue in 1667. Comparisons of other copies have detected similar instances of corrections of marginal misnumberings in other places, with sometimes also such a discrepancy in the text as a with instead of in. How are these small variations to be accounted for? Not, I think, by the supposition that, for the later issues, a leaf was occasionally cancelled and reprinted on account of some error discovered in it—in which case copies of the later issues should be found the most correct, whereas we have just seen an instance to the contrary. Rather, I think, by the supposition that, during the original slow printing by hand-press of the whole first impression of 1,300 or 1,500 copies in 1667, that happened which sometimes happens even yet in a printing-office with steam machinery: viz. the detection of errors in time to correct them for a portion of the impression. Thus several of the sheets, as kept in bales for binding, might be in different states of correctness, and a later-bound copy might have one of these sheets in its first or less correct state.
