The following discussion at the July 2011 Dickens Symposium at St. Anselm's College in Manchester, New Hampshire, opened with these initial remarks by Philip Allingham about Dickens's knowledge of precisely where the month's illustrations would be inserted when the novel wrapped up its serial run:
1. Now, as to how serial readers would know where to position their monthly illustrations, the answer is simple. Until they received the nineteenth monthly number, with its chapter and illustration indexes, they would simply have to guess, except of course for the illustrations for the last double-number. I suspect that readers collecting the monthly instalments would have attempted to position the illustrations themselves, then would have corrected these positionings once they could consult the index of plates, whose page numbers would have been received as "facing" such-and-such a page. Then there is the matter of whether a serial reader would have kept and had bound in the 32 monthly pages of advertisements ("The Advertiser"); I suspect not because their presence would have doubled the size of the volume for binding. Consequently, rather like flyers in today's newspapers, these advertisers were probably discarded each month.
Elizabeth Cayzer has already made these pertinent points about how Dickens serial would have been assembled at the conclusion of the nineteen-month run:
When assembled as a monthly part, the pictures were tipped into the episode ahead of the text. Such a juxtaposing of two plates would easily set up a train of reverberations in the reader's mind. 
In the closing double number, the reader would find two more illustrations, the Frontispiece and the Title-page. Placed here they allow him to reflect upon what he has read. Later, placed at the front of a bound copy of the novel, they also announce the main themes and concerns of the book. 
The monthly reader from the first, then, is engaged critically as he or she must determine precisely where each of the two full-page, loose illustrations will fit when the novel's parts are bound. The weekly reader, on the other hand, did not have the illustrations to remind him of previous actions or anticipate developments in the plot.
2. Response from Prof. Gareth Cordery, Emeritus, University of Canterbury, New Zealand:
A few further observations on the positioning of illustrations. Their placement together at he beginning of monthly numbers encouraged readers to view them as a pair. Thus for the eighth number (December 1849) of David Copperfield Phiz pairs "I Make the Acquaintance of Miss Mowcher" with "Martha," suggesting a connection through hair (Mowcher is a hairdresser hired by Steerforth, Martha's long, loose hair indicates her fallen nature) and hence Steerforth's later seduction of the about-to-fall Little Em'ly.
Robert Patten, in his masterful entries in the Oxford Reader's Companion to Dickens, writes about the positioning of the illustrations in bound editions. "Instructions were also provided to binders about where to insert the plates, facing that page most applicable to the image" (294), and in the final double monthly number there were "directions to the binder about where to place the illustrations if the owner of the parts wants to reassemble them into a volume" (527). But he does not say who issued these instructions or directions so our questions about who (Dickens, Phiz, publisher or a combination of these) was finally responsible for the positioning of the illustrations in a bound edition remain. A similar question about who decided which advertisements were to be included in the Dickens Advertisers, those pages of ads that sandwiched the letter press for each monthly number, and their placement, was raised at the recent Dickens Symposium in Manchester (NH), but likewise remains unanswered.
You speculate that readers probably discarded the "Advertisers" when having the parts bound. In what follows Patten does not confirm this but he does offer extra information about what readers did or could do with those monthly numbers and how printers took advantage of this burgeoning market:
"And because serial parts were hard to track and to care for over nineteen months, many customers bought the novel in parts as it came out, and also at its conclusion bought a second copy encased in hard covers to keep permanently. Alternatively, publishers offered various kinds of bindings for the parts, and did a lively business supplying missing parts before they were reassembled as volumes. .
Finally, the illustrations were positioned at times outside the novel — in shop windows, quite separate from the monthly wrapper. Like the wrappers, they were powerful marketing tools, encouraging both the literate and illiterate to examine them for clues about that particular installment.
The wrapper to Dombey & Son [click on thumbnail to enlarge image.]
Your comment that wrappers invited readers to speculate on future events reminds me of CD's comments to Foster (6 September 1846) that the wrapper to Dombey & Son perhaps had "a little too much in it.” In other words, he felt that wrappers might reveal too much, thus endangering the mystery of what happens next, which was so important to keeping his readers hooked and willing to shell out their shilling for the next number. This particular wrapper shows how much of the plot and characters CD had already sketched out, but on the other hand too much specificity might discourage flexibility and inhibit CD's famed (and in my opinion desirable) tendency to head off in any direction his imagination might lead him. In any case was the overly specific wrapper for Dombey & Son the reason for the overly vague (in my opinion) wrapper to his next novel, David Copperfield? Or was it related to its autobiographical nature where he revisits (reluctantly?) his painful past? We will never know, but wrappers clearly exploited an important aspect of the reading experience, namely anticipation. There is one interesting exception, the 1846 edition of Oliver Twist, a reissue in ten separate monthly parts of the original serialisation in Bentley’s Miscellany, 1837-39. Its wrapper, designed by Cruikshank, includes several rectangular vignettes of his original illustrations (some reworked) and a few new ones. Here satisfaction replaces anticipation, as readers identified scenes and characters made famous eight years earlier. A rare case, then, of a wrapper functioning retrospectively rather than prospectively.
- Taking Off the Wrapper: David Copperfield Anticipated, May 1849.
- Hardy, Dickens, Serialisation, and Illustration: A Rebuttal of Alan S. Watts' “Why Wasn't Great Expectations Illustrated?"
- A Note on Phiz's Wrapper Design for Dickens's A Tale of Two Cities (1859) in Monthly Serialisation
- Phiz's Cover for the MOnthly Parts of Bleak House
- Marcus Stone's August 1864 Wrapper for Our Mutual Friend
Cayzer, Elizabeth. "Dickens and His Late Illustrators. A Change in Style: Phiz and A Tale of Two Cities." Dickensian 86, 3 (Autumn, 1990): 30-4.
Patten, Robert L. "Illustrators and Book Illustration." The Oxford Reader's Companion to Dickens. Oxford University Press, 1999.
Last modified 11 August 2011