[begin page 571]

After nearly five years of intermittent work on A Connecticut Yankee in King Arthur’s Court, Mark Twain completed the manuscript in the spring of 1889. He turned it over to his publishers in April. The process of transforming it into a published book took eight months more. During that time the book changed in many ways: there are hundreds of substantive differences between the manuscript and the first American edition, ranging from one-word substitutions to five-hundred-word additions and eight-hundred-word cuts; and there are some twelve hundred differences between the accidentals of the manuscript—its punctuation, spelling, and so on—and those of the book.1 These changes were made on typescripts and proofs which were thrown away after they had served their purpose. They were made by Mark Twain, by the typist who transcribed his manuscript, by the compositors who set the book in type, and by the printer’s and publisher’s proofreaders—who played a far less restricted role than their modern counterparts, functioning much as copy editors do today. The editorial problem is thus a historical problem. To discriminate between the changes Mark Twain made and the alteration others imposed on his work requires the imaginative reconstruction of lost documents—the typescripts, proofs, and instructions which passed between the author’s study, his publisher’s office, and the print shop during those eight months. To determine Mark Twain’s intentions we must trace the movement of those documents and recreate the activity of the people through whose hands they passed.

A Connecticut Yankee was the first book that Mark Twain had submitted to his publishing house, Charles L. Webster and Company, [begin page 572] since Adventures of Huckleberry Finn, which had appeared in 1885. The company was depending on the new book to restore its failing financial health. For Fred J. Hall, who had succeeded Webster as manager of the firm in February 1888, it was the first time he would have to take responsibility for a book by the house’s senior partner. Understandably anxious about handling the assignment, Hall made a great show of consulting Clemens about every aspect of the book’s production and marketing, and it is chiefly in his letters to the author that the record of the book’s transmission from manuscript to type is preserved.2

The record begins with the Webster company’s terse report to Mark Twain on 16 April 1889 that two copies of the book had been typed up.3 The absence of an acknowledgment from the firm that the manuscript had been received suggests that Clemens delivered it to his publisher personally. He still had the manuscript at the beginning of the month, for he read from chapter 28 at the Hamersley home in Hartford on April 1. Probably he relinquished it when he visited New York on April 8 and 9 to speak at a dinner for a touring American baseball team.4 That he put in a portion of his time in the city at the Webster company’s office is confirmed by a letter from the publisher [begin page 573] suggesting that Ellen Mackay Hutchinson, the literary editor of the Tribune, be given the title of the book and “the substance of what you wrote when you were here to be incorporated in our descriptive Circulars” so that she could “make a very brief mention of it not more than fifteen or twenty lines” in her paper.5

The book remained at the publishers in New York until at least the second week in May. On May 8 Hall wrote to tell Clemens, “I have just about completed reading the type written copy we have in the office, of your book ‘The Yankee In The Court Of King Arthur.’ ” He went on to praise the book as “the best book you have ever written,” to compare it with the work of Swift and Cervantes to those authors’ disadvantage, and to confess, “It makes me feel nervous as to my ability to properly place it on the market and give it the circulation it deserves.“6

In spite of this lavish praise, however, Hall felt nervous about more than his own ability to secure a large sale: something about the book made him uneasy. He decided to seek a second opinion of the work, and began the delicate task of maneuvering Mark Twain into accepting the criticism of an outside reader. The man Hall chose was Edmund Clarence Stedman, a well-known poet and literary critic whose eleven-volume Library of American Literature (coedited with Hutchinson) the Webster company was publishing. When Clemens visited New York in early May, Hall began his campaign; on May 14, he returned to it in a carefully disingenuous letter:

The next day Stedman unwittingly revealed Hall’s strategy by writing to the author on his own behalf.

[begin page 574]

In speaking of “works as individual & original as yours,” Stedman may have been echoing Hall and disclosing the grounds of the publisher’s anxiety. In any case, his letter gave Hall a more immediate reason to worry—Clemens might realize that he had been telling each man that having Stedman read the work was the other’s idea. As soon as he heard about Stedman’s letter, Hall hastened to disavow his role:

In the event, Hall need not have been concerned. Clemens replied to Stedman, “I shall be most heartily obliged to you. By the time you get back from your triplet I shall have the MS back in Mr. Hall’s hands for you.”9

Clemens had retrieved the typescript of A Connecticut Yankee on a trip to New York in early May.10 It is impossible to pinpoint when he returned it to Hall, but presumably he kept it until late June: Stedman mentioned on July 7 that he had received the book the week before. Clemens must have turned the typescript over to his publisher in person, for on July 5 Hall referred to “two conversations” he and [begin page 575] Clemens had recently had regarding “points which Stedman should keep in mind while reading it.”11

The interval of nearly two months in May and June was the longest period Mark Twain had with the typescript, and was relatively free of interruptions, as well. The typescript was heavily revised, and although he had later opportunities to make changes in it, it was undoubtedly at this time that it received its most thoroughgoing revision. With the delivery of the revised typescript to Stedman, events began to move rapidly, and Mark Twain’s concern shifted from perfecting his book to producing it, as he sifted the final candidates for illustrator, worked to arrange the book’s publication in England, and became absorbed in the details of planning the typesetting, printing, and marketing of the American edition.

It took Stedman only two days to read the book, and by July 5 he had returned the typescript to Hall. In a note he gave the anxious publisher his heartening verdict, calling the book “a great and unique work,” and adding in conversation that “he saw scarcely any thing to change in it.” He accompanied the typescript with some suggestions for Mark Twain, urging that eight or nine scattered passages be changed.12 What he objected to in every case was a particularly hyperbolic exclamation or description, like the reference to the hermit who was so dirty that “you couldn’t tell the sewer from the saint in the dark.” On July 7 Stedman wrote Clemens directly, including in the letter the suggestions he had handed to Hall. He had found little to quarrel with, but although he told Mark Twain, “You have written a great book: in some respect your most original, most imaginative,” an undercurrent of apprehension accompanied his approval. He was afraid that others would not understand the book as he did: “Whether the ordinary critical reader will take in its real claims to importance, is a serious question.” He extolled the “riotous,” “original,” and “imaginative” and criticized the “superlative,” exaggerated, and extravagant: what he commended lies dangerously close to what he condemned, and sometimes the two co-exist uneasily in his praise: “Of course, when you let yourself loose, ’tis somewhat like a stallion just out of the [begin page 576] paddock. But ’tis remarkable how finished, and in what good taste, your whole work—considering the theme & its possibilities—is. There is scarcely anything which I wished to change.” Finally, he was uncertain about the reception the book would have: “I suppose the sale of this unusual book will depend somewhat on the ‘working capacity’ of your firm. But it will make a great noise, at all events.”13

However much these warnings and the doubts latent in his letter qualified Stedman’s approval, he did endorse the book. Hall sent one of the typescripts back to Mark Twain and began to consult with the printers, the firm of Jenkins and McCowan. It took the author some time to respond to Stedman’s suggestions, however. After Theodore Crane’s funeral, he left his family in Elmira with Crane’s widow, Olivia Clemens’ foster sister, and removed to the Hartford house. In Hartford he continued to revise the book, but he seems not to have made the alterations called for in Stedman’s letter before interrupting his work and sending the typescript back to New York, in preparation for a visit from Howells.14 On July 17 Hall wrote him, “We will have the two type written copies compared so that the corrections you have made will be the same,” and asked, “Do you not want one of these sets in order to make use of Mr. Stedman’s corrections?” Two days later he wrote again, with increasing urgency, “We have had the manuscript compared so that the corrections are the same in both copies. If you desire to have any one revise the manuscript more, or to make any further corrections in it yourself it would be well to do so as soon as possible as we want to get to work setting it up very soon.” As he completed this second appeal he received a wire from Mark Twain responding to his first and added in a postscript “have sent mss.”15

When he received the typescript Mark Twain promptly rewrote every passage to which Stedman had objected. He later explained to Howells,

[begin page 577]

While he sacrificed his “darlings,” Mark Twain continued the general revision of the typescript, added the preface (which is dated July 21), and inserted at least one new blast of opinion. An entry in his note-book shows that at the end of July he added the section on royal grants to chapter 25 while he was choosing excerpts from the book for publication in the Century magazine.17 On July 29 Hall wrote Mark Twain, “As soon as Manuscript is received, the printers are ready to jump right to work.” His letter crossed with a wire and letter from Mark Twain, which Hall acknowledged the next day: “We . . . note you say you have forwarded the book by express, but it has not as yet been received, undoubtedly will be some time during the day.”18 When Mark Twain next saw his book, it was in type.

Once Stedman had returned his opinion of the book and Daniel C. Beard had been chosen to illustrate it, Hall and his assistant, E. H. Rosenquest, began preparing the typescript for the press. In mid-July Rosenquest made a careful estimate of the book’s size and had the printers set up a sample page. Hall was confident that he had allowed ample time to produce the book for the Christmas market; he was urgent in his pleas that Mark Twain complete his revisions and relinquish the typescript, he said, only because “it is a dull time now with our printers and they can work better on it than they can later.”19 On [begin page 578] July 23 Hall visited Beard, armed with Clemens’ instructions to have the artist consult with Stedman and with his views “about the quality of the pictures, that is, to have more or less humor in some of them, but not too much.”20 He returned from the meeting in panic. In the longest letter in the company files for this time, Hall confessed, “I am afraid I have made a serious blunder in the calculation of the time necessary to get out your book,” and concluded, “The whole trouble lies in my failing to recognize the difference between elaborate, detailed drawings such as Mr. Beard is going to make, and the sketchy, outlined drawings such as Kemble who illustrated Huckleberry Finn made. One takes three or four times longer to make than the other.” Hall had hoped to issue the book by mid-November, but he found that Beard could not submit his last drawings before November 1. Although he took the blame on himself, Hall subtly reminded Mark Twain about his own role in creating the problem, by echoing the author’s instructions about the illustrations: “This time I want pictures, not black-board outlines & charcoal sketches. If Kemble illustrations for my last book were handed me today, . . . I would put them promptly in the fire.” And he openly told his employer that his desire to keep close control over the pictures created a bottleneck: “One very important item is the time consumed in sending the proof to you at Elmira and especially in sending the drawings to you for approval before they are put into the engraver’s hands.”21

Hall could offer only one way out of the dilemma—an extremely tight production schedule in which “the entire process of book making can be going on at the same time, from the binding down to the illustrating.” In outlining his proposal, he provided an extraordinarily detailed, and illuminating, picture of the way the book was to come into being.

Even as Hall was writing his letter, the crisis was easing. Mark Twain reconsidered his injunction “to have more or less humor . . . but not too much” in the pictures and dashed off a note of his own: “Upon reflection—this: tell Beard to obey his own inspirations, & when he sees a picture in his mind, put that picture on paper, be it humorous or be it serious. . . . They will be better pictures than if I mixed in & tried to give him points on his own trade.”23 Thus at a stroke he eliminated one of Hall’s great fears, that the time it took Mark Twain to approve the drawings would become a major source of delay. Clemens’ reply to Hall’s long letter does not survive, but it appears to have been equally reassuring. He evidently approved Hall’s plan, for Hall replied jauntily, “We will keep all parties well drummed up,”24 and later letters show that the scheme was indeed put into effect. Clemens’ letter apparently also relieved the pressure somewhat by authorizing a short delay in publication: on July 29, Hall named [begin page 580] December 10 as publication day in a letter to Mark Twain’s English publisher; and in early August both Clemens and Hall urged the Century magazine to change its schedule and include excerpts from the book in the November rather than the December issue, so that the magazine would provide advance publicity.25 Clemens also had a specific suggestion for minimizing the time spent reading proof. He seized on Hall’s insistence that “it will be necessary to have . . . the proofs read and returned as soon as possible” as an opportunity to enforce a requirement of his own: his longstanding, and futile, wish to have the compositors and proofreaders follow his copy exactly. No doubt he pointed out, as he had to each of his previous publishers, that if no changes were imposed he could concentrate on reading for sense without worrying that his careful punctuation had been tampered with, or pausing to restore it. Hall promised: “Now in regard to reading proof of the book, I will see that it is compared with the greatest of care and made to conform exactly with the manuscript. As fast as the page-proofs are perfected we can send them to you and any changes necessary to be made in the plates can then be done.” Finally, Mark Twain seems to have worried that the pressure on Beard would compromise the quality of his work. Hall replied, “Mr. Beard says that the hurry will not affect his work except, in this way, that he will not be able to put as much detail into the small pictures. . . . However, he will make a good job of it.”26

The new timetable also meant that Mark Twain could not put as much detail into his proof revisions as he had in the past. When the Webster company had published Huckleberry Finn, he had had two opportunities to read proof, and he had altered both galleys and pages. The schedule for A Connecticut Yankee restricted him to one reading, and the fact that he read pages rather than galleys limited the length of the corrections he could make—too long a cut or addition would force the printers to remake all the pages in a chapter. Hall’s decision to have plates made before the author returned his proofs reinforced this prohibition and also inhibited even short revisions (although nineteenth-century printers altered plates more willingly [begin page 581] than their later counterparts, and the changes they made were more extensive than later practice allowed). Once the plates were corrected, only the most immediate second thoughts could cause them to be altered again before they were imposed and printed under the accelerated schedule.

Nevertheless, Mark Twain had no intention of neglecting the proofs: he knew they provided his last opportunity to change his mind, and always took them seriously. With the book in press, he marshaled his forces, writing Howells on August 5, “Mrs. Clemens will not listen to reason, or argument; or supplication: I’ve got to get you to read the book. . . . But you will not have to take it at a bite. I will spread it thin, & leave resting-spells all along. The proofs, thoroughly corrected, & then revised & re-corrected, shall go to you as re-revises, from time to time, from the office in New York.”27 Mrs. Clemens had an eye ailment that prevented her from attending to the proofs, but it is hard to imagine that if her eyes had been sound Mark Twain would not have found another reason to ask Howells for help. The pattern was well established. Howells had read and helped revise Tom Sawyer, The Prince and the Pauper, and Huckleberry Finn before they were published; and Mark Twain ordered the Connecticut Yankee proofs sent to Howells before he had agreed to read them—probably, in fact, on the same day that he asked Howells to undertake the chore.28 Placing the burden of the request on Mrs. Clemens simply warded off the sense of incurred obligation that might otherwise have grown up between the two men. As Clemens had anticipated, Howells replied with his usual kindness and generosity, “You know it will be purely a pleasure to me to read your proofs. . . . If I didn’t want to read the book for its own sake or your sake, I should still want to do it for Mrs. Clemens’s.”29

The proofs of the first pictures arrived in mid-August,30 followed [begin page 582] shortly by word that the first galley proofs were being pulled by the compositors, and distressing news: “With reference to reading proof of your book, the punctuation as it is in the Manuscript is different entirely from the punctuation as our proof reader conceives it ought to be.” Anticipating Mark Twain’s reaction, Hall added, “I told him plainly, that he must have no opinion whatever regarding the punctuation, that he was simply to make himself into a machine and follow the copy. Is this all right?” Mark Twain’s reply to Hall was temperate: “You are perfectly right. The proof-reader must follow my punctuation absolutely. I will not allow even the slightest departure from it.” He gave freer play to his feelings in response to an inquiry from Howells about when to expect the proofs: “They’ll be along very soon, now, I guess; for yesterday Mr. Hall wrote that the printer’s proof-reader was improving my punctuation for me, & I telegraphed orders to have him shot without giving him time to pray.”31

Clemens’ proofs did come along soon, and they probably didn’t cool his temper any. He had told Howells that they would get “rerevises”—that is the proof would be read orally against the typescript, the compositor would make the marked corrections and pull a new proof that would be checked against the first, and further corrections would be made if necessary, before the author’s proof was pulled. Without errors to distract them, Mark Twain and Howells could concentrate on substantive revision. Hall had agreed to this procedure, but he was preparing to go to Montreal to negotiate the contract for the Canadian edition and had turned matters over to Rosenquest. On August 21 Rosenquest wrote, “We mail you today the first batch of page proofs. Our proof reader is now revising these [begin page 583] with the galley proofs and we will send you a revised set as soon as ready.”32 Mark Twain’s usual reaction to uncorrected proofs was an explosion of fury, but when the Connecticut Yankee proofs arrived in the morning mail on August 24, he seems to have greeted them with professional resignation. When he informed Howells of their arrival, he did remark parenthetically that they were “not revises,” but he passed quickly to expressing his delight in the pictures, which he was seeing within the text for the first time. The proofs covered at least the first seven chapters, and perhaps the first nine.33 Mark Twain set to work on them at once, without waiting for the corrected proofs, and read so rapidly that he was finished by evening. He must simply have closed his eyes to errors and departures from copy, trusting the proofreader to catch and correct them, for he returned the proofs so quickly that Rosenquest was able to tell him, “The corrections you marked in the proofs have all been made in the pages before they were cast.”34 Along with the proofs, Clemens returned a drawing of the book’s cover that Rosenquest had sent him so that he could approve Beard’s plan to shorten the title on the cover and spine by dropping the word “Connecticut.” Apparently he did remonstrate about the state of the proofs, for when Hall returned he told Clemens that he would send “correct proofs” henceforward; and he must have complained again about the compositors’ disregard for his punctuation, for Rosenquest assured him that they had “taken out the words ‘which laid the other one out cold’ without altering the punctuation,” and Hall added, “We note what you say about proof reading, and have again cautioned the reader to make no changes.”35

By the time Hall returned, Mark Twain had a new complaint: [begin page 584] Howells had received no proofs. From the tone of Hall’s letter of explanation, it would appear that this time Clemens had erupted.

Rosenquest’s error shows that Hall’s plan to engage simultaneously in all the stages of manufacture had really been put into effect. By “proofs” Rosenquest had understood Clemens to mean proofs taken after the type was made up into pages, but before the pages were plated or imposed. The type pages were placed in galley trays in consecutive order, probably in groups of three, and then proofs of them were printed on galley-length paper. By “sheets” Rosenquest had understood Clemens to mean proofs taken after the type pages had been plated and imposed. Proofs of these plated pages were then printed on a large sheet, which when folded and cut would present them in their proper order.37

Mark Twain bowed to Hall’s explanation of the error: “I made the confusion myself, then. Well, I might have expected it. I apologize.” With the managing partner back in the office, an orderly flow of proof began, but it took Howells, who had gone to Ohio to visit his father, two weeks more to begin reading. When he did, he was enthusiastic—“It’s charming, original, wonderful”—but he also had a few suggestions for changes: “So far I find nothing but a word or two even to question.”38 There is no way to tell what changes in the text stem from Howells’ criticism, but if the words he found to question in his first session with the proofs were in the first hundred pages or so, they are [begin page 585] probably still in the book. By the time Howells began to read, a second batch of proofs had been sent out, and the pages in the first batch had probably already been printed. The proofreading of A Connecticut Yankee was off to a bad start: as a result of Rosenquest’s misunderstanding, Mark Twain never read corrected proofs for nearly one-fifth of the book, and Howells saw them too late to have an effect. By September 22 Howells had passed chapter 13,39 but he may still have been too far behind to make any alterations in the text. Rosenquest acknowledged the return of Mark Twain’s second set of proofs on September 23; by now production was in full swing, and he told Clemens, “These corrections will be made in the plates before any copies are printed from them.” He added, “We will send you some more page proofs tomorrow.”40

With a smooth routine established, proof flowed steadily between New York and Hartford for the next month and a half. Little external evidence remains to mark its movement or disclose the quality of Mark Twain’s alterations. On October 16 Beard was working on illustrations for chapter 32; he pointed out that it would be anachronistic to have the Marcos serve turkey at their dinner. Mark Twain had just returned proof for chapter 27, exercising his veto over the illustrations for the only time recorded in the correspondence. Howells was fully caught up, for he wrote the next day, “This last batch, about the King’s and Boss’s adventures, is all good.” Although the pace was rapid, Mark Twain was fretting about “tedious & hurtful delays” because he feared that Howells would not have the completed book in hand in time to review it for the January issue of Harper’s Magazine. Hall was more cheerful; he reported, “Mr. Beard is keeping right up to time and we do not anticipate any delay.” Four days later, on October 29, he announced in a postscript to a letter devoted to finances, “The Yankee is all in type.”41

[begin page 586]

Once the printing and proofreading were well underway, the door-to-door selling of the book began. On October 10 Hall reported, “We expect prospectuses of your book up to-day or to-morrow certainly, and will commence the canvass at once. The general agents have already sent out a great many circulars and made many appointments and are waiting for the prospectuses to pitch right in.”42 The prospectus was the chief weapon in the subscription publisher’s sales campaign. Each salesman carried one of these slim volumes; it contained the table of contents and samples of the text and illustrations from the book he was selling, incorporated specimen bindings, and included blank ruled pages to record the names and addresses of subscribers. It was Hall who chose the contents of the Connecticut Yankee prospectus. His selections emphasized, as the advertisement bound in the volume said, that the book was “a keen and powerful satire on English nobility and royalty, a book that appeals to all true Americans,” and he followed Mark Twain’s instructions to “be careful not to get any religious matter in.”43 The prospectus was almost certainly set independently of the book. It includes portions of chapters 43 and 44, which the book’s compositors had not yet reached, and portions of “A Word of Explanation” and chapters 1, 2, 6, and 7, which had already been plated. The physical differences between the prospectus pages and the book pages are marked. Only one initial letter—the first—is included in the prospectus; no other picture is inserted in the text. Two pages of text conclude with tailpieces; the rest of the pages are either devoted exclusively to an illustration or are of solid type. Consequently, many of the prospectus pages do not correspond to pages of the book, and more than half the individual lines vary as well. Probably Hall handed the compositors a list of typescript page numbers, and [begin page 587] they composed and paged the prospectus on the spot. The accidentals of the prospectus vary from both the manuscript and the book, and such errors as “snu” for “sun” and “recking” for “reeking” suggest that it was proofread only cursorily or not at all.

By October 14, Hall could boast:

Instead of waiting to use up the first printing of the prospectus, Clemens and Hall decided to add a gathering of pictures to the unbound signatures that remained:

[begin page 588]

By October 29, when Hall announced the completion of the typesetting, the canvassing was going so well that he was also able to tell Clemens, “We have increased the edition of your book ten thousand and are manufacturing twenty-five thousand now.”46

Howells finished his reading of the proofs on November 9 and wrote the next day to congratulate his friend. He did urge at least one more change, in the speech Clarence makes when he impersonates the commander of the knights’ army in chapter 43. Mark Twain balked at Howells’ suggestion, and poor Hall, who had declared that “complete sheets of the book will be ready by the 15th,” had to watch the days go by as proof passed through him to Stedman, to whom Clemens referred the question, back to Clemens, who capitulated, and finally to the printer on or soon after November 14. The last change was made, and the final forms went to press; the first impression was completed no later than November 20, when Hall sent Clemens a set of stitched sheets.47

Mark Twain had one last trial in store for Hall. In July he had made notes correlating the historical sources he had consulted with the sections of A Connecticut Yankee he had used them for. He planned to note his sources in an appendix, to demonstrate that he had not invented or exaggerated the cruelty his book portrayed. With the manuscript he had included another note reminding himself to make an appendix “in support of the assertion that there were no real gentlemen & ladies before our century.”48 Ultimately he abandoned his plans for an appendix, but the idea of verifying his book’s accuracy persisted, and on November 24, with A Connecticut Yankee at the bindery, it suddenly burst forth in a new form. He ordered Hall to print an American edition of one of his source books, George Standring’s People’s History of the English Aristocracy, a radical Englishman’s racy exposé of the aristocracy’s corruption and sexual intrigue.

Unfortunately, the letter in which Hall talked Clemens out of this scheme does not survive. But it is not hard to imagine how his heart must have sunk at the idea of “piling on the printers,” to say nothing of the hint that Mark Twain might be thinking of delaying his own book’s publication. The brainstorm had passed by December 3, when Hall wrote to tell Clemens that arrangements had been made to insure that newspapers receiving advance copies of A Connecticut Yankee would not jeopardize English copyright by printing reviews before December 8. The Webster company published the first American edition on schedule on December 10.49

Hall’s plan to illustrate, typeset, plate, and print the American edition all at once proved to be an efficient way to meet the Webster company’s obligation to issue its edition in early December. But this schedule nearly cost Mark Twain his English copyright and in fact forced his English publisher, Chatto and Windus, to produce a flawed edition from early proofs which was never brought up to date. At the same time, a letter of violent protest from Mark Twain, or rather his biographer’s misinterpretation of it, has left generations believing that Mark Twain was able to publish his book in England at all only through a combative insistence on his right of free expression coupled with a judicious compromise with censorship at the last.

Although the correspondence in the Mark Twain Papers and the data gathered in preparing this edition are sufficient to provide an outline of the facts surrounding the publication of the English edition, the full story has only recently come to light, in Dennis Welland’s account of the relations between Mark Twain and his English publisher and in the documents from the Chatto and Windus files which his study makes available for the first time.50

In January 1887 Andrew Chatto read a newspaper report of Mark Twain’s Governor’s Island reading and inquired about the book: “I lose no time in writing to ask . . . if you are sufficiently advanced with it to enable me to make any announcement or preparation in connection with the issue of an English edition.” Replying that the [begin page 590] book was still in its infancy, Clemens said he planned “to work on it again next summer” and promised, “It will be submitted to you first of all foreign publishers.” Chatto prodded Clemens in August of 1887 and again in July of 1888; to the second letter, Clemens responded, “We go home to Hartford a week hence; & if at that time I find I am two-thirds done, I mean to try to persuade myself to do that other third before spring.” Although he did indeed complete his manuscript by spring, no one told Chatto, who did not discover the fact until his annual inquiry about the book, this time in a letter to the Webster company, elicited the information.51

At about the time that Stedman was reading the typescript of A Connecticut Yankee, Chatto again sought a commitment from Clemens and the Webster company to make concrete plans for the book’s British publication. In a letter enclosing royalties from the sale of Mark Twain’s earlier books, he wrote:

As Welland says, “Mark Twain’s reply to this straightforward and goodnatured request is one of the most remarkable and least explicable letters in his entire correspondence with Andrew Chatto.”52

Clemens wrote the letter from Hartford, but sent it through his publishing house, where a typed copy was prepared before the original was transmitted to Chatto. Confronted with this typescript in the files of the Mark Twain Papers, in isolation from Clemens’ other correspondence with the firm, Albert Bigelow Paine quite naturally concluded that Mark Twain’s letter was a reply to a lost letter from Chatto and Windus. When he published Clemens’ letter in Mark Twain: A Biography, Paine introduced it with his deduction that “the London publishers of the Yankee were keenly anxious to revise the text for their English readers.” Five years later, in Mark Twain’s Letters, the [begin page 592] hypothetical letter from Chatto and Windus achieved solidity: “Mark Twain’s publishers,” Paine wrote, “asked that the story be especially edited for the English edition. Clemens, however, would not listen to any suggestions of the sort.”54 In fact, Chatto had never addressed any such suggestions to Clemens; Chatto had never even seen the book, much less objected to its satire against “monarchy and its several natural props.” But Paine’s plausible guess hardened into “fact” and led students of Mark Twain to believe that Chatto and Windus had tried to censor (or did censor) Mark Twain’s work, a belief that persisted for more than sixty years while the files of the publishing house remained inaccessible.

Mark Twain’s letter to Chatto was not written in response to a demand for changes in his text, but in anticipation of one. Chatto had never made such a demand before, and few changes of the sort have been found in the authorized English editions of Mark Twain’s books: only John Camden Hotten, who pirated editions of Mark Twain’s work in the early 1870s, regularly invented “modifications of his language . . . to fit it for the sensitive English palate.”55 Yet it is not entirely surprising that Mark Twain expected Chatto to insist on further editing. However little basis there was for Clemens’ conviction that his work had been censored in the past, he did genuinely believe it. A year earlier he had recorded the accusation in his notebook: “Every time in a book I happen to speak of a king differently from the way one speaks of God, or of a noble differently from the way one speaks of the Son of God, it is stricken out of the European reprints.— [begin page 593] Seems to give the proofreader the dry gripes.”56 Moreover, in writing A Connecticut Yankee he had written a controversial book, and one that had as one of its central themes a denunciation of British political institutions. But if it was not unreasonable for Mark Twain to expect his book to be challenged, it surely was unreasonable for him to write so brutal and accusatory a letter.

We can only speculate about why he felt so thin-skinned that he adopted the vehement rhetoric that misled Paine into thinking his letter to Chatto was a counterattack. Part of the answer may lie in his situation: he was deeply affected by Theodore Crane’s lingering death and by his family’s grief. He was alone in the Hartford house, and lonely. In a letter to his daughter, Susy, he depicted his mood on the day he wrote to Chatto:

Perhaps, too, something of the uneasiness that Hall and Stedman felt about the book had touched Mark Twain. He had recently finished his second reading of the typescript; Stedman’s letter had reached him within the week. At the very least, reading Stedman’s letter with its prediction that “Some blasted fool will surely jump up & say that Cervantes polished off ‘chivalry’, centuries ago” and its warning that “there will be various rows & rumpuses” would have confirmed Mark Twain in his belief that the book was going to provoke criticism and controversy. And in addition to the anxiety, the letter may have awakened a kind of combative pride: after all, he had subjected his work to extensive scrutiny. Mark Twain’s insistence that he had “taken laborious pains to so trim this book of offence” that Chatto would “not lack the nerve to print it just as it stands” is insulting; but stripped of its rhetoric, his claim to have taken especial care in revising the book is true. What he told Chatto in a vituperative, [begin page 594] finger-pointing fashion, he would later tell Howells with friendly assurance, when he predicted that Howells would find little to question in the proofs because with Stedman’s help he had “rooted out . . . blasts of opinion which are so strongly worded as to repel instead of persuade.”58

Typically, as soon as Mark Twain had vented his anger, he relented. When he revised his typescript in late July for the third and last time, he did a complete about-face and instructed Hall to omit the second paragraph of the preface, with its comments on the divine right of kings, from the copy sent to England and to authorize Chatto to make further cuts if he wished in the text of his edition. He singled out the section on royal grants, which he had just added to the printer’s copy, as an example of a passage Chatto might omit. Hall acknowledged Clemens’ instructions on July 30 and passed them on to Chatto on August 5.

Before receiving this letter, Chatto had replied to Clemens’ first: “I am very pleased to learn that the Yankee at the Court of King Arthur is to be published on December 10th next. I do not think there is any possibility of your writing anything I should not be pleased to publish.” Apart from using the truncated preface, Chatto did not avail himself of Mark Twain’s offer to alter the text. Probably it was just as well that he didn’t, for on the day the book was published, the New [begin page 595] York Times printed an interview with Clemens which virtually repeated the accusations, and some of the language, of his July 16 letter to Chatto.

Clemens had given the interview to Robert Donald, an English journalist, early in December. Donald arranged to have it published in the Pall Mall Gazette and then placed a somewhat different version with the Times.60 It is impossible to say whether Clemens actually believed what he told Donald, or whether the interview was puffery, calculated to pique the curiosity of the English while appealing (as most of the publicity for the book had) to the patriotic and populist sentiments of Americans. When Chatto read the interview in the Gazette, he finally remonstrated with Clemens. Enclosing the clipping, he pointed out,

In fact, although Chatto did not cut out or alter a word, the English edition of the book is not identical to the American. To secure English copyright, British law required that the English edition be published before the American edition. Imperial copyright, to protect the book in Canada, was also tied to this requirement, along with the additional [begin page 596] requirement that the author be on Canadian soil the day the book was published. Yet Clemens seems never to have grasped the principle that he had to allow sufficient time for the text and plates of the pictures to reach England from the United States and that Chatto needed time thereafter to set type and then to print. Through the years, Chatto’s letters are filled with pleas to get plates and copy to him, and Clemens’ and his publishers’ are filled with blithe unconcern. In the case of A Connecticut Yankee, the problem was exacerbated by the tight schedule Hall had devised. Hall recognized as early as July 24, when he first informed Clemens that he had miscalculated the time needed to prepare the book, that the English edition might cause trouble.

Hall wrote Chatto on July 29. From Chatto’s response, it would appear that Hall sent him a virtual duplicate of his July 24 letter to Clemens:

Chatto’s reply also shows that he was mindful of past problems and delays, and acutely aware of the cavalier attitude Mark Twain’s publishers had shown to them.

If Chatto had known that when Hall wrote him he had not yet received the finally revised printer’s copy from Mark Twain, and that it would be three more weeks before the first proofs could be expected from the printers, even he might have lost his remarkable composure.

Hall wrote Chatto on August 19, when the galleys began to arrive, promising to begin sending proofs and electroplates at once. Shipment waited until after the pages had been made up, however. Probably the first shipment to Chatto corresponded to the first set of proofs sent to Mark Twain, which were dispatched on August 21. The package that arrived in London on September 16, with less than three months remaining before publication, consisted of the first 113 pages; it included proofs of the pictures, but no plates for them.64 Moreover, the proofs did not represent “corrected copy of the text”: they were probably the same uncorrected proofs that Mark Twain had received; they certainly did not include the author’s proof revisions. In chapter 2, for instance, Mark Twain changed Hank’s question to Clarence to read, “What is the name of that apparition that brought me here?”; but the English edition retains the tamer manuscript reading, “What is the name of that man who brought me here?” Sending early proofs to England was not an unusual practice, but ordinarily they were followed by revised copies to allow the English publisher to correct his text. On September 5 Clemens told Hall, “Send what you think best to Chatto. I should think that after my corrections have been made, the matter is ready for him, whether it go in slips or in sheets.” Then he showed just how loose his grasp was on the problems facing Chatto by asking, “Why should he need a set of our sheets for copyright purposes. Won’t his own printed and bound book answer?”65 Hall at least understood enough to ignore Clemens and continue to send unrevised [begin page 598] proof, as the twenty-five or so manuscript readings which Mark Twain altered in the Webster proof but which remain in the Chatto and Windus edition show.66 Among the choicer changes that failed to get into the English edition were the substitution of “boarding-house butter” for “condemned butter,” and the addition of the question Hank asks Morgan le Fay when she is torturing her prisoner: “Where is the profit?” Equally often, however, language that Mark Twain softened in the Webster proofs remains in the English edition: when Hank learns that he must undertake the quest with Sandy he is “as glad as a person is when he is scalped” in the American edition, but “as glad as a person is when he is disembowelled” in the English, for instance. And the change Howells insisted on in chapter 43 was never sent across the Atlantic.

In spite of Hall’s corner cutting, by mid-November Chatto began to apprehend disaster. “We are disappointed at the slowness with which proofs and blocks are reaching us,” he told Hall; “we ought to have the whole of the book and illustrations in our hands by this date, in order to publish by the date originally fixed. Even then, the time would have been short for so heavy a book. Please to send us everything outstanding at once.” By then Hall had already sent the last proofs; he could hardly have done so earlier, for the typesetting had been completed only two weeks before Chatto wrote. By November 22 Chatto had received the last of the proofs and could promise, “Our printers will have it all in type if Messr Webster let us have the Electros of the remaining illustrations without delay.—we are anxiously awaiting their arrival.” He also pleaded with Clemens, “Do not publish the American edition before we have received a perfect copy of the book that we can register.” When Hall sent Chatto’s letter on to Clemens, he assured him airily, “The matter spoken of in this letter has been made all right. That is, they have received, long ago, electros of cuts, and also six complete sets of sheets.” Hall was wrong, as he almost learned to his [begin page 599] sorrow. The sheets did not reach England until the day before publication, as Chatto disclosed when he wrote on December 6, with a pardonable pride that was probably altogether lost on Mark Twain:

The Canadian edition was produced after a comedy of errors. Through much of the summer, Hall and Clemens assumed that it would be published by Samuel Dawson, who had published Mark Twain’s previous books. In fact, Dawson had retired, as they discovered in mid-August. When Hall wrote Chatto on August 5, suggesting that he omit the royal grants section from his edition and seek permission for further cuts, he realized that the changes might jeopardize the Canadian edition. In the uncertain state of copyright law, Hall believed that the Canadian edition might have to be identical to the English, and told Clemens, “Dawson cannot use our plates if there are to be any material changes in the English edition of the book.” He suggested that the Canadians produce a pro forma copy by putting their title page on a set of American sheets, then “have Chatto & Windus have duplicate plates made of their Edition and forward one to Dawson.” Clemens apparently missed Hall’s point about copyright when he replied, for Hall acknowledged, “The best way would be to have him use our plates . . . whether he objects to them or not . . . to secure copyright, then if he wants to order a set of plates from Chatto & Windus to print his regular copies from, all right.”68

[begin page 600]

When Hall finally wrote to Dawson, Dawson recommended the Rose Publishing Company to replace him and suggested, “Rose could set the book up in type and print it without illustrations so as not to delay us, and then . . . get a set of plates from England.” Hall then wrote Clemens, “In accordance with your suggestion, I will see Dawson and talk the matter over with him, and after getting full instructions as to copy right law, will go from there to Rose and make an arrangement for the Canadian Copy.” This decision took Hall to Montreal at the crucial moment in late August when the first proofs were being sent to Mark Twain, thus producing the most damaging side effect of the assumption that the British publishers would insist on alterations in the book. Hall returned with a contract with Rose, now in the Mark Twain Papers, which called for Rose to use a set of English plates. But immediately upon his return Hall told Clemens, “We do not need to send proof to Canada, as Rose is to use our plates.” On November 21 Hall instructed Clemens “to cross the Canadian line, at any point, and register at some hotel in Canada and remain there during the hours of publication in England.” At Clemens’ request he asked Rose to confirm these instructions, and when Rose did so he was able to add, “Our book is now off press and in the bindery.” The author made the required trip on December 6, and the next day Hall was able to assure him, “The copyright . . . is all right.” Chatto and Windus had verified “that they had published . . . the day that you were in Canada.”69 The Canadian edition, printed from a duplicate set of Webster company plates, was published on December 8.

Almost from the moment he began to write A Connecticut Yankee, Mark Twain contemplated publishing excerpts from it in a magazine. He had experimented for the first time with this method of attracting public notice to a new work when he printed selections from Huckleberry Finn in the Century magazine in 1884 and 1885. In February 1886, a week after he told his nephew, Charles Webster, “I have begun a book, whose scene is laid far back in the twilight of tradition . . . & got myself into the swing of the work,” he added a postscript to a letter to Clarence Buel, an assistant editor of the Century, suggesting, “When [begin page 601] I get this book done, I think a chapter or two will read very well in the Century.”70 The matter was left there until the spring or summer of 1889, when Mark Twain promised the magazine’s editor, Richard Watson Gilder, a portion of A Connecticut Yankee in exchange for an article he had withdrawn from the magazine. In late July, Robert Underwood Johnson, the Century’s associate editor, reminded Clemens of his promise: “I have just been up to Marion on a visit to Mr. Gilder and I found him very desirous to have from you, for the Nov. number if possible, the promised part of your new book ‘A Conn. Yankee at the Court of King Arthur.’ ”71

The Century’s editing of Huckleberry Finn has become something of a cause célèbre in the critical debate over the emasculating effect of “the genteel tradition” on American literature. Many critics have assumed that as a matter of course any work submitted to the magazine was pruned by its editors of improprieties of diction, and that A Connecticut Yankee was no exception. But although there are almost sixty substantive variants between the Century printing and the first American edition, few of them can be attributed to an editorial impulse to soften language or sentiment. As in the case of the American and English editions, production of the article had to be completed in a great hurry, and the pressure of time inhibited searching editorial alterations. But more importantly, Mark Twain was intimately involved in the preparation of the Century excerpts. He selected the portions to be published, had a new typescript made from the Webster company typescript to serve as the magazine’s printer’s copy, revised the Century typescript, supplied the bridge passages that link the various excerpts together, carefully read the proofs to approve or disapprove of editorial modifications, and even supplied captions for some of the illustrations.72

[begin page 602]

When Johnson’s letter arrived, Mark Twain was in the midst of his last revision of the Webster typescript. He set to work at once to make selections from it for the magazine. He began rather haphazardly, jotting down in the middle of a list of source books—which he was keeping in his notebook for his planned appendix—the note “204 the typescript page number 2d—beginning of Century article.” The entry refers to a portion of the episode in chapter 24 where the monks from the Valley of Holiness take their first bath. On reflection, Mark Twain decided not to use this chapter in the magazine, doubtless because to include it would have violated his policy of avoiding references to the church or to politics in publicity for the book.73 A few pages later in the notebook, Mark Twain made one more random entry—“For Century, ch. 39—revolver.”—before setting down an orderly list of the excerpts that did in fact appear in the magazine.

[begin page 603]

Presumably he transmitted these instructions to the Webster company when he returned the typescript; on July 30 Hall promised “to have the type-writing for the Century matter attended to.”75

Johnson wrote to Clemens again on August 2 and asked whether Kemble would be doing the illustrations for the “ ‘Yankee at the Court of Cyrus the Persian or Peter the Great or Benjamin the Unready’—or whoever it is.” Aware that time was short, he added, “We have the article down ‘for December sure’ and only want to know whether we can expect a shy at the pictures too.” Clemens replied the next day, identifying Beard as the artist and urging that the excerpts appear in the November issue, as Johnson had originally proposed. He suggested that Johnson “drop in & discuss with Mr. Hall. If needful, the Beards would doubtless skip over to that part of the book & make those pictures at once.” Hall agreed that “publication in the Century in November would be better than in December, as it would help the sale of the book previous to its publication.”76 Johnson met with Hall on August 7, and they struck the final bargain. “It is agreed between Mr Hall and yourself and myself,” Johnson wrote, “that the chapter is to go into the November number.” Time was so pressing that Johnson simply walked off with “Beard’s first drawing—the armed figure charging upon the dry goods drummer in the tree” and “sent it to be processed, having little margin of time for illustration.” (When he later received Mark Twain’s selections and found that the frontispiece was not related to any of them, he appealed to the author, who saved the picture by supplying the caption “The Yankee’s Reception in Arthurdom.”)77

Mark Twain also realized that time was short, and even before he learned that the November date had been agreed to he told Hall to send the new typescript of excerpts to him in Elmira. A worried Hall wrote on August 8, “I have received your letter of August 6th. . . . You speak therein of not having received the type-written copies of the selections for the Century. These were mailed to you three or four days ago, and if you have not received them by this time, let me know and I will have duplicates made.” Clemens had received the typescript by the time Hall wrote, either in a later delivery on August 6 or on the [begin page 604] seventh. Once again he showed how rapidly he could work, reviewing the typescript and writing the bridge passages within a day, and notifying. the magazine, “I’ve done as you required—done my very levelest best to get it to you in time for the November number—& I reckon I’ve succeeded.—Hope so, anyway. I mail it to-night.”78

Although he had only a very short time with the Century typescript, Mark Twain used it to tinker once again with his text. Changes like the addition of “—and I didn’t thin out the rats any for his accommodation” to the description of Merlin’s cell in chapter 7, and the substitution of “I reckon they added on a couple of days” for “probably the facts would have modified it” at the end of the same chapter, could hardly have been made by anyone but Mark Twain. He clearly felt an irresistible urge to play with his language whenever the opportunity presented itself. In addition to these literary changes, he also made local revisions specifically for the magazine. Some of these, like the identification of Clarence as “the page,” provided information that the magazine reader would need because he didn’t have the whole book before him. Other local revisions continued the process of removing religious or political material: Mark Twain cut out a number of references to the church and deleted mention of “my West Point—my military academy” in chapter 10, and swore the Century staff to secrecy about these matters. “Mum’s the word about the political bearing of the book,” Johnson promised him.79

When they received the typescript, the Century editors did make some changes in the text. It was probably they who deleted “Raphael was a bird” from Hank’s complaint about the quality of the artwork in Arthur’s castle, and changed “upsetting” to “turning over” a sentence later. They were careful to see that no infinitives were split, that subject and verb agreed in number, and that prepositions and pronouns were properly used. But what lit their editorial fire—and set Mark Twain’s temper ablaze—was not the text but the title and introduction. On August 15 Johnson wrote Clemens that he had “sent the MS to the printer,” but he added, “It will not do to use the title you have given us for the extracts from the new book.” To persuade Clemens to change it, he enclosed a letter from Gilder:

[begin page 605]

But Clemens was not reasonable this time. On receipt of Johnson’s and Gilder’s letters, he fired off a telegram withdrawing the article from the magazine. He insisted on making it absolutely clear to the journal’s readers that they were seeing only a small portion of a long work. Either in the title or in the introduction he had apparently referred to the excerpts as “scraps,” and he remained immovable about identifying them in this or a similar way. On August 16 Gilder wrote, “I have just gone to town to work . . . & the first thing I see is your telegram.” He wired in return, “A second withdrawal would be serious injury to our plans. We can arrange title to suit you.” In the letter that followed, he sought to mollify Clemens further while continuing to press for a change.

[begin page 606]

Clemens grumbled, but he agreed to meet the magazine halfway. On the envelope of Gilder’s letter, he wrote:

Johnson replied for the magazine in what appears to be a cover letter for the proofs.

When the November Century appeared, it carried Mark Twain’s piece under the title of the book, with no subtitle, but with an introduction that began, “Here follow a few incidents from an impending book of mine which bears the above title.” Hall congratulated Mark Twain on his victory:

Mark Twain worked on the magazine and book versions independently. He never had the two typescripts in his possession together, and he never had proof of the two versions at the same time. He managed to duplicate one of the revisions he made in the Century typescript—the substitution of “boarding-house butter” for “condemned butter”—in the first American edition, but none of his other literary changes for the Century found their way into the Webster edition. The existence of these two independently revised texts places [begin page 607] the editor of a critical edition in a curious dilemma. Are the Century revisions to be regarded as changes Mark Twain made only for the magazine, or do they represent his latest intention, which was unfulfilled in the book? It would be difficult to fault an editor who concluded that the Century variants should be excluded from his text: Mark Twain had his last opportunity to revise when he read the Webster proofs, and he did not enter the magazine changes there. Nevertheless, his failure to do so does not necessarily mean that he wanted to confine them to the magazine. The Century typescript presented Mark Twain with his last opportunity to make literary changes freely, without the constraints imposed by having his work in type. By August 6 or 7, when he received the Century typescript from Hall and revised it, the Webster company typescript was already at the print shop being set. Since the Century revisions are, for the most part, small refinements and adjustments of language, Mark Twain may have felt that it was wiser not to call for resetting; by the time he got the Webster proofs, he may even have forgotten about them. But an editor is bound by no such limitations, and may legitimately ask what the author would have wished to do under ideal circumstances, in this case if he could have set the Century and Webster typescripts side-by-side. It seems clear that about a dozen of the changes Mark Twain made in the Century were made for no reason except that the author thought they clarified or improved his text. These changes have been adopted in this edition.85

Apart from the Century excerpts, only two authoritative texts of A Connecticut Yankee survive: the manuscript and the first American edition. Although he authorized the shortened preface, Mark Twain made no changes in the text of the English edition; beyond contracting for their publication, he had nothing to do with the first Canadian edition, which was printed from a duplicate set of Webster company [begin page 608] plates, or with the Continental edition, which was typeset from a copy of the English edition and published by Bernhard Tauchnitz in Germany early in 1890.86 Nor, collation shows, did the author alter the text for any of the later editions published in his lifetime: the second and third English published by Chatto and Windus in 1892 and 1897, the second American published by Harper and Brothers in 1896, and the third American published by the American Publishing Company in 1899 and reissued by the American Publishing Company and by Harper and Brothers under several names thereafter.87

Since Mark Twain read no proof for the prospectus or for the first English edition, they have no textual authority. They do, however, have textual significance. By preserving the readings of their setting copy—the Webster typescript and unrevised Webster proof, respectively—they help toward a reconstruction of those lost stages and allow certain small variants that would otherwise remain problematic to be assigned confidently to the typist, to the compositors, or to Mark Twain’s revising hand. Because it was set from a typescript made from the Webster typescript, the Century magazine, too, helps in this process of reconstruction.

If the prospectus or the Century varies from the manuscript and shares the reading of the first American edition, we can be certain that the change to the first American edition’s reading occurred on the Webster typescript. Similarly, the agreement of the first English and first American editions against the manuscript rules out the possibility [begin page 609] of a proof change: the shared reading must be either the reading of the typescript or the result of a compositor’s error. If the prospectus or the Century agrees with the manuscript against the first American edition, such agreement preserves the reading of the typescript and demonstrates that the change to the first American edition’s reading occurred either during the typesetting or during the proofreading. Finally, the agreement of the first English edition with the manuscript preserves the reading of the uncorrected Webster proofs and thus pinpoints the change as an alteration in the American copy of the proofs. Knowing the stage at which a change occurred does not always by itself allow one to say whether Mark Twain made it; but it does narrow the range of speculation. It is pointless to notice, for instance, that the manuscript is easily misread, when a change was made during the typesetting or proofreading and the printer had only the typescript to consult; it is idle to think about the lay of the compositor’s case, but reasonable to visualize a typewriter keyboard, when a change occurred in the typescript.

Comparison of the four texts typeset from lost authoritative copy shows that Mark Twain made very few changes in proof. His time with the proof was severely limited, whereas he worked on the typescript on three separate occasions—in May and June, in mid-July, and in late July—for a total of about two months. And the typescript allowed him scope: even without the evidence of the Century, prospectus, and first English edition, it would be clear that all of his longer revisions, the deletions from and additions to the text as he had left it in the manuscript, could have been made only in the typescript. Not only did the inflexibility of the production schedule and the intractability of type metal prevent major revisions in page proof, but having clean, typed copy to revise was still new to Mark Twain—an opportunity to treasure and to use. Six years earlier, he had been the first author to employ a typist to produce printer’s copy: “previously,” as he said, “books had been copied for the press with the pen exclusively.” That first typescript was of portions of Life on the Mississippi; portions of Huckleberry Finn were also typed; A Connecticut Yankee was the first of Mark Twain’s books to be typed from beginning to end. In a testimonial to the man who typed Life on the Mississippi, Mark Twain explained what the change from “books copied for the press with the pen” meant to him.

[begin page 610]

There were also problems inherent in having the manuscript typed, however. Just as compositors introduced errors when they set their copy in type, so the typist introduced errors when preparing the typescript: the fresh opportunity for revision of the text also provided a fresh opportunity for its corruption. The most common scribal mistakes are misreading the author’s hand; memorial error, which occurs when the copyist attempts to hold too much of the text in his head and to set it down without reference back to his copy; and eye skip, which occurs when the copyist takes his eye off the copy, then returns to the same word farther along on the page, leaving out a portion of a sentence. These errors are common to all transcription; to them must be added mistakes peculiar to typing, the chief among them being the familiar error of hitting the wrong key. Furthermore, the typescript was not hedged about by the safeguards established in the print shop: no team of proofreaders compared the typescript to its copy to weed out mistakes, and when a typist fortuitously produced a sentence that made sense, though it didn’t follow its copy, the change was likely to remain undetected. The typist who transcribed A Connecticut Yankee seems to have been particularly prone to skipping over repeated phrases, altering Hank’s speaking style without altering the meaning of the narrative.

In addition to providing a new stage at which errors could be introduced, typing the manuscript opened a new avenue for impromptu editing. Like the Webster company proofreader who found Mark [begin page 611] Twain’s punctuation “different entirely from the punctuation as . . . it ought to be,” the typist of A Connecticut Yankee appears to have volunteered “corrections” and “improvements” freely, both of punctuation and of language. This tendency to edit the text of the course of typing it was so marked that it provoked Mark Twain into adding a sarcastic commentary on the typist’s literary ability to the book in May, when he read the typescript for the first time. From it we can gather what sorts of changes the typist felt free to make. In chapter 25, when one of the knights who applies for a lieutenancy in the new standing army is unable to answer The Boss’s questions, Hank bursts out:

Critical editions seek to strip this accretion of error and uninvited alteration contributed by typists, compositors, and proofreaders from an author’s work, leaving a text that is as close as the evidence permits to the text the author wanted. Modern editorial theory stipulates that to produce an authoritative scholarly edition an editor must place before his readers not only such a text, but the evidence and reasoning by which he has constructed it. Whenever the editor must make a choice, he is obligated to report and defend it. As a first step toward fulfilling this obligation, the editor designates a “copy-text”—the basis for his own text which he must follow in all particulars except where emendation is required. The choice of copy-text determines the form in which the evidence for establishing the final text is organized; and it informs the reader of the source of every reading in the edited text, for each reading must either originate in the copy-text or be listed with its source as an emendation. From the evidence presented in the textual apparatus, the reader must be able to reconstruct the copy-text and to judge the reasonableness of each decision to incorporate a variant reading in the edited text or to exclude it. Central to the theory of copy-text is W. W. Greg’s distinction between substantives and ac- [begin page 612] cidentals. Greg recognized that authors are less likely to revise their accidentals than to revise the wording of their work, whereas copyists, proofreaders, and house editors are likely to alter the formal features of a work while, for the most part, respecting its substantives. This house styling is likely to prevail, since the economics of publishing put great pressure on authors to acquiesce in it. In general, therefore, the theory of copy-text proposes that an editor adopt as copy-text the text closest to the author’s hand, incorporating as the author’s revisions most of the substantive variants appearing in later authoritative texts, while, on the whole, adhering to the copy-text’s accidentals.89 The theory fits Mark Twain’s practice remarkably well.

The copy-text for this edition of A Connecticut Yankee in King Arthur’s Court is Mark Twain’s manuscript. (The Globe edition of Sir Thomas Malory’s Morte Darthur serves as copy-text for the extended quotations from Malory in “A Word of Explanation” and in chapters 3, 9, 19, and 42.)90 The author’s literary revisions in the Century magazine and his revisions in the first American edition have been [begin page 613] adopted as emendations and are recorded, along with a few simple corrections supplied by the first American edition, or by the editor himself, in the textual apparatus. The emendations list thus becomes a window on Mark Twain’s workshop, through which we can see the changes he made, chiefly on the typescript, as he shaped his work into final form. The vast majority of the substantive variants in the first American edition—the cuts and additions and the major changes in diction and dialect—can be ascribed to no one but Mark Twain: their authenticity is guaranteed by their intrinsic character. Something over one hundred and fifty substantive variants in the first American edition are rejected from the text of this edition, and the copy-text is restored. They are, for the most part, readily identifiable as transcription errors, or as the typist’s, compositor’s, or proofreader’s “uninvited emendations” or sophistications.

There are, however, a substantial number of variant readings less easily accounted for, and the treatment of these readings of indeterminate origin constitutes a difficult editorial problem. There is no way to determine with any approach to certainty, for instance, whether Hank should call Clarence “nothing but a work of imagination,” as he does in the manuscript at 83.7, or “nothing but a work of the imagination,” as he does in the first American edition. Similarly, either the typist or Mark Twain could have added the second “I” to Merlin’s declaration, “I can, fair sir, and I will” at 104.3; and the author, typist, compositor, or proofreader could have changed “are” to “is” in the clause “I don’t sit down and wait for doctors when Sandy or the child are sick” at 446.27–28. When a variant is entirely indifferent, and we have no evidence beyond its existence to guide us, the reading of the first American edition has been rejected from the present text. Examples include “work of the imagination” mentioned above, and such readings as toward/towards, henceforth/thenceforth, those/these, and in/of. Although Mark Twain could certainly have made some changes that alter neither the meaning of his text nor his mode of expression, there is no way to distinguish his changes from those of others: by restoring the manuscript we at least preserve what is certainly an authorial reading and insure against admitting what may very well be an unauthorized change. Variants are also treated conservatively when some factor or combination of factors—the ease of misreading the manuscript, the likelihood of sophistication, or the [begin page 614] probability of transcription error, for example—tends to throw suspicion on the printed text. The change in number of the verb in the phrase “Sandy and the child are sick,” for instance, holds Mark Twain to a stricter grammatical account than his usual practice warrants. Although, once again, he was perfectly capable of making such a change, it is so typical of the efforts of a transcriber or proofreader to correct the work at hand that the manuscript reading has been restored, as it has been when split infinitives are corrected or when words are transposed or skipped over and no evidence can be marshaled to show that the change was not a transcription error.

To reject small and relatively indifferent changes wholesale, however, would be to deny to Mark Twain the two months’ effort he put into polishing the typescript. With so much time to tinker with his prose, there can be little doubt that he made many small changes in wording, as he did on preserved typescripts of later works. Therefore when some evidence, however slender, can be adduced to suggest that Mark Twain made a change in the typescript, the reading of the first American edition has been adopted as an emendation. Changes in the first American edition that are analogous to changes Mark Twain made while revising his manuscript (the expansion of contractions, for instance); linked changes as in the alteration of “the three knights” and “the three damsels” to “these three knights” and “these three damsels” at 181.10; changes that follow a pattern throughout the book, like the deletion or alteration of the pronoun “that”; and otherwise indifferent variants in heavily revised passages are taken to be Mark Twain’s work. So are changes in dialect, to which Mark Twain was always sensitive.

There remains a class of problematic variants that cannot be characterized in any of these ways. Though small, they are not indifferent, yet we have no evidence beyond the fact of variation to guide us in choosing between the manuscript and the first American edition. Lacking any evidence of who made the changes, we must try to discern from their character why they were made. Three examples will have to suffice. When Merlin says, “I can fair sir and I will,” the addition of the second “I” makes his pronouncement still more solemn, and thus heightens the contrast between his style and Hank’s breezy colloquialism. In chapter 10, Hank declares, “Everybody could be any kind of a Christian he wanted to” (127.13–14); the manuscript reads [begin page 615] “any kind of Christian he wanted to.” The article was added in the typescript, and memorial error is a possible explanation, but the locution is typical of Mark Twain. Lastly, the substitution of an article for a personal pronoun in the description of St. Stylite running a sewing machine—“the motive power had taken to standing on one leg . . . there was something the matter with the substituted for “his” other one” (260.27–29)—completes the dehumanization of the hermit, reducing his leg to a part in the engine that he has become. Although in each of these cases and in others like them the later reading has been adopted as Mark Twain’s improvement of the text, none of these judgments can claim to be definitive. To rely on the copy-text, however, would risk repealing Mark Twain’s revision of the typescript; while trying to be cautious, we have chosen instead to exercise our experience of Mark Twain’s habits and style and make our decisions on literary grounds.91

Reliance on the copy-text is the sure way to preserve Mark Twain’s punctuation, however. In spite of his pleading, in spite of his orders, in spite of his threats, both the typist and the workers at the print shop lightened and normalized his punctuation. Mark Twain punctuated so that his sentences would be heard, not just read. It makes a real difference if the cadences of a sentence like this one in the manuscript are altered: “Hammering each other—for we stepped aside and looked on while they rolled and struggled, and gouged, and pounded and bit, with the strict and wordless attention to business of so many bulldogs” (384.32–35). The first American edition added a comma to follow each verb in the sequence so that it reads, “Hammering each other—for we stepped aside and looked on while they rolled, and struggled, and gouged, and pounded, and bit, with the strict and wordless attention to business of so many bulldogs.”

The various correctors imposed a thousand or more alterations on Mark Twain’s punctuation. In recovering the texture of Mark Twain’s own pointing, we inevitably lose the bit of revision he may have done on the typescript and in proof, but since these few alterations cannot [begin page 616] be disentangled from the many supplied by other hands, and since restoring the manuscript always preserves an authorial reading even if it occasionally sacrifices a later one, the manuscript’s punctuation has been respected. The first American edition frequently dropped the manuscript’s punctuation where Mark Twain substituted a word or phrase in the typescript for his original wording in the manuscript. At 364.1, for example, “soul” in the first American edition is Mark Twain’s revision of “spirit” in the manuscript; “spirit” is followed by a comma in the manuscript, but there is no punctuation after “soul” in the first American edition. Mark Twain may not have rewritten his punctuation when he interlined his revision, leaving the typescript open to easy misinterpretation; or the omission might well be house styling. The copy-text punctuation has therefore been restored in such instances.

Only a few mechanical corrections, such as the addition of closing quotation marks at the end of a speech, have been made. Such technical errors as the run-on sentence “A second rope was unslung, in a moment another slave was dangling” (424.3–4) have not been emended: the construction is typical of Mark Twain, and the use of a comma instead of a period or semicolon adds to the sense of urgency in the narrative. Other idiosyncratic but typical constructions include the comma before the verb in such sentences as “The most conspicuously situated lady in that massed flower-bed of feminine show and finery, inclined her head” (69.17–19), and the final comma in the questions “Well, then, what are they waiting, for?” (which the first American edition preserved, at 169.8) and “What did you want him to do that, for?” (which the first American edition dropped, at 203.14–15). This latter use of the comma seems to be an attempt to reinforce the question mark—to find a graphic equivalent for a voice rising with puzzlement.

Precisely because Mark Twain’s punctuation was largely oral, changes in emphasis marks provide the major exception to the rule that Mark Twain’s manuscript usage has been scrupulously followed. It was his consistent practice to supply a new set of exclamation points and italics during revision, and these features of the text have therefore been adopted from the first American edition.

Mark Twain was less careful about the other accidental features of his text than about his punctuation. Although he was a good speller, he [begin page 617] misspelled some words habitually—like “sieze” for “seize”—and some carelessly. He was also inconsistent at times about spelling, compounding, and capitalization. He used “ancle” and “ankle,” spelled Guenever’s name two ways and Sagramour’s three, and spoke of “knight-errantry” and “knight errantry,” and of “the Church” and “the church,” for example. Correcting errors of this sort was properly within the province of the typist, compositor, and proofreader.92 Where they succeeded, this edition follows the corrections in the first American edition; where they failed, this edition attempts to extend their efforts, correcting such misspellings as “phazed,” for example, and imposing consistency in spelling, compounding, and capitalization, so long as the author’s inconsistency appears to be unintentional and without purpose.93 Of course, it is considerably less easy to detect and correct inconsistent usages than to correct outright spelling errors. When Mark Twain writes “armor” twenty-seven times and “armour” once, and when that once is in the phrase “Sagramour’s arms and [begin page 618] armour,” so that the our from the name carries over, one’s choice is easy enough. At times, the manuscript guides us to Mark Twain’s preference, as when he alters “Sagramor” to “Sagramour,” “offense” to “offence,” or “the church” to “the Church.” But when he writes “ice cream” and “ice-cream” once each, the choice must be arbitrary.94

All of the emendation discussed above is recorded in the list of emendations in this volume. A few mechanical changes have been made without notation in the emendations list. Forms peculiar to the written page have not been transferred to the printed page. The ampersands that Mark Twain used throughout all his manuscripts have been expanded to “and”;95 “&c.” is expanded to “etc.” Superscript letters are lowered to the line. Where Mark Twain designated chapter headings by writing “Chap.” and leaving the number to be filled in, his abbreviation is silently expanded and followed by an arabic numeral. In a few instances, Mark Twain used a short dash after a period at the end of a line in the manuscript. These dashes are clearly distinguishable from his longer marks of punctuation, and appear to be a holdover from the old convention of filling out a short line with a dash to avoid confusing the end of the line with the end of a paragraph. These short dashes have been dropped.96 Punctuation following italic words is styled italic according to the usual practice, whether or not Mark Twain underlined the punctuation. Mechanical errors in inscription occasioned by incomplete revision in the manuscript are [begin page 619] recorded in the list of alterations in the manuscript, to avoid overburdening the emendations list. They include double punctuation, when Mark Twain inserted new punctuation without deleting the old; a capital letter in the middle of a sentence, when he added new matter to the beginning of a sentence without changing the capital of the word that originally began the sentence to a lowercase letter; and a period in the middle of a sentence, when he added matter to the end of a sentence without deleting the original period. The use of small capital letters following the initial letter that begins each chapter is an editorial convention. A number of the illustrations have been reduced slightly in size to accommodate them to the page width of this edition. Finally, although Mark Twain did not write the table of contents, chapter titles,97 list of illustrations, or captions for the illustrations, all are included in this edition. They are styled to accord with the text of this edition.

B.L.S.

[begin page 620]

Description of Texts identifies and discusses editions published in Mark Twain’s lifetime and specifies copies collated and examined in the preparation of this edition.

Textual Notes specify those features of the text discussed generally in the textual introduction, record all of Mark Twain’s marginalia in the manuscript, and discuss adopted readings and aspects of Mark Twain’s revision which require fuller explanation.

Emendations of the Copy-Text lists every departure from the copy-text and records the source of the reading in the present text. It is in two parts: substantives and accidentals. The list of accidentals includes the reading adopted when a compound word is hyphenated at the end of a line in the copy-text.

Rejected Substantives records all substantive readings in the first American edition which are rejected from the present text, and all substantive readings in the prospectus, the Century, and the first English edition which vary from the first American edition.

Alterations in the Manuscript provides a record of every revision that the author made in the manuscript.

Word Division in This Volume lists ambiguous compounds hyphenated at the end of a line in this volume, and gives their correct form for quotation.

[begin page 621]

The following texts have been collated, and the collation results are reported in the textual apparatus because of the light they shed on the writing and revision of A Connecticut Yankee. The symbols on the left are those used in the apparatus to identify the texts. Of the symbols, only Cent, A, and I-C appear in the emendation lists; all the symbols are used in the list of rejected substantives.

MS     

Mark Twain’s manuscript of A Connecticut Yankee in King Arthur’s Court, in the Henry W. and Albert A. Berg Collection of The New York Public Library.

The manuscript consists of 934 pages on five varieties of paper. Mark Twain wrote in three distinct colors of ink and added some revision in pencil. The manuscript leaves are numbered on one side only and, except for occasional insertions and additions on versos, are inscribed on one side only. Probably for his own convenience in working with a manuscript of such length, Mark Twain numbered his manuscript in two sequences. Howard G. Baetzhold’s designations for these sections, MS I and MS II, are retained here. The manuscript is complete, although Mark Twain added pages to and removed pages from the numbered sequences; MS I is numbered 1–482 and MS II is numbered 1–400. Pages that Mark Twain removed from the manuscript before it was given to the typist are in the Mark Twain Papers. They are transcribed in full in the list of alterations in the manuscript.

Centered headings in the list of alterations in the manuscript identify the sections of the manuscript by paper type and ink color. The five varieties of paper are designated as follows:

[begin page 622]

Keystone Linen (1) is a cream-colored, laid stationery measuring 22.4 by 14 centimeters (8 13/16 by 5½ inches). It has vertical chainlines 2.4 centimeters (15/16 inch) apart and is watermarked with the words “Keystone Linen” and a keystone-shaped device containing the initials “JCB.” The leaves were torn from a tablet with a red binding, fragments of which are visible on the top edge of some leaves.

Light green paper is a laid stationery measuring 22.5 by 14 centimeters (8⅞ by 5½ inches). It has horizontal chain-lines 2.7 centimeters (1 1/16 inches) apart and is otherwise unwatermarked. The leaves were torn from a tablet with a red binding, fragments of which are visible on the top edge of some leaves.

Keystone Linen (2) is a cream-colored, laid stationery measuring 22.2 by 14.1 centimeters (8¾ by 5 9/16 inches). It has horizontal chain-lines 1.8 centimeters (¾ inch) apart and is watermarked with a picture of a globe rising out of an envelope labeled with the words “TRADE MARK.” The leaves were torn from a tablet.

Blairs Keystone is a light blue, laid stationery measuring 22.5 by 13.9 centimeters (8⅞ by 5½ inches). It has horizontal chain-lines 2.7 centimeters (1 1/16 inches) apart and is watermarked with the words “Blairs Keystone.” The leaves were torn from a tablet.

The manuscript contains a single, torn half-sheet of cream-colored, wove, unwatermarked paper measuring 21.4 by 13.8 centimeters (8 7/16 by 5 7/16 inches).

Three ink colors are distinguishable in the manuscript: blue-black, black, and blue. Much of the black ink has faded with age to brown. The list of alterations in the manuscript gives a full account of writing materials, with centered headings identifying the three large sections of the manuscript in which each ink color predominates. There are revisions in pencil throughout the manuscript.

Malory     

Sir Thomas Malory, Morte Darthur, the Globe edition, edited by Sir Edward Strachey. London: Macmillan and Company, 1870–1886.

Strachey’s index, from which Mark Twain borrowed names, was added in 1869. Mark Twain could have used any of the impressions dated 1876, 1879, 1883, 1884, or 1886; the one he did use has not been identified.

Mark Twain used the following passages from Malory. They are keyed both to the present edition and to the Globe edition.

48.27–50.15	How . . . host.	Book 6, chapter 11
72.10–74.7	“Right . . . did.”	Book 1, chapter 23
121.11–122.21	Then . . . not.	Book 7, chapter 28

[begin page 623]

173.14–180.4	So they . . . age—”	Book 4, part of chapter 16, chapter 17, parts of chapter 18
221.14–223.3	“Now . . . grace.*	Book 4, parts of chapters 24 and 25
462.21–463.12	Then . . . oft-times.	Book 21, part of chapter 4

Cent     

“A Connecticut Yankee in King Arthur’s Court,” Century Magazine 39 (November 1889): 74–83.

The excerpts printed in the Century correspond to the following passages in this edition:

50.23–37	I am . . . saying.
98.1–105.32	Inasmuch . . . back so.
126.25–130.9	I was . . . flavor.
428.1–439.19	Home . . . left.

The Century article prints the frontispiece and the illustrations that appear in the present text on pages 47, 68, 432, and 434. The top part of the illustration on page 51 was used as a tailpiece for the magazine article.

Several explanatory paragraphs used in the Century to provide context for the passages from A Connecticut Yankee are printed in Appendix F of this edition.

Pr     

Prospectus. A Connecticut Yankee in King Arthur’s Court. New York: Charles L. Webster and Company, 1889.

The prospectus consists of eleven pages of front matter, sixty-eight pages of text and illustrations, two pages of advertising for A Connecticut Yankee and a special gathering of sixteen pages of illustrations (see Appendix E), and thirty-two pages lined into columns for the recording of subscribers’ names and addresses.

The excerpts printed in the prospectus correspond to the following passages in this edition:

45 title–27	Preface . . . 1889.
47 title–48.31	A Connecticut . . . When his
50.16–53.4	As . . . penmanship Pr page ends with “pen-”
56 title–57.36	CHAPTER . . . always
60 title–63.19	CHAPTER . . . didn’t
86.35–92.2	the argument . . . saw I
98 title–100.31	CHAPTER . . . poor
140.24–142.28	I was . . . said—
209.3–210.10	A master . . . it was

[begin page 624]

210.31–212.16	So I . . . presence, and
226 title–228.1	CHAPTER . . . snag or
284 title–286.8	CHAPTER . . . Aldermen, who
288.2–288.33	Men . . . qualifications
294.6–296.20	“I yield . . . parents.
329.23–334.35	But . . . death-rattle.
396 title–398.7	CHAPTER . . . fazed.
431.17–436.4	Down . . . re-arranged. I
442 title–444.22	CHAPTER . . . couldn’t
483.23–484.15	“What is that?” . . . awful
484.17–486.10	These . . . and had
488 title–493.18	CHAPTER . . . it.

In addition to the illustrations printed in the special gathering, the illustrations on the following pages were used in the prospectus, although they were not placed or ordered as they are in the first American edition: pages 47, 53, 55, 62, 64, 81, 85, 89, 93, 99, 112, 117, 123, 137, 148, 432, and 434.

The illustration that follows appeared in the prospectus but was not used in the book.

X

In the prospectus this illustration served as a tailpiece following the end of chapter 29 (p. 334), although it bears no relation to the text at that point. It was evidently replaced in the book by the more elaborate drawing of the same scene on page 58.

A     

First American edition. A Connecticut Yankee in King Arthur’s Court. New York: Charles L. Webster and Company, 1889–1891. BAL 3429.

[begin page 625]

Collation revealed only one variant among the impressions of this edition: at 71.9 the reading “too modest too glorify” was corrected to “too modest to glorify” in the 1890 and 1891 impressions.

E     

First English edition. A Yankee at the Court of King Arthur. London: Chatto and Windus, 1889–1896. Discussed in BAL entry 3429.

The following editions were found to be derivative and without authority.

A Yankee at the Court of King Arthur:

Leipzig: Bernhard Tauchnitz, 1890 ( BAL 3648); set from the first English edition.

London: Chatto and Windus, 1893; an unillustrated two-shilling edition set from the first English edition.1

A Connecticut Yankee in King Arthur’s Court:

Toronto: G. M. Rose and Sons, 1889; and Rose Publishing Company, 1890; printed from a duplicate set of Webster company plates of the first American edition.

New York: Harper and Brothers, 1896; set from the first American edition.

Hartford: American Publishing Company, 1899 ( BAL 3456), and New York:

Harper and Brothers, 1904 ( BAL 3670); set from the 1896 Harper edition.

The following sight and machine collations were performed at the Mark Twain Papers in the course of preparing this edition.2

[begin page 626]

A photocopy of the manuscript was read against the first American edition three times,3 and the manuscript itself was completely checked against the printer’s copy for this edition (a marked Xerox reproduction of a copy of the first American edition) and against the tables of the textual apparatus. The photocopy of the manuscript was read once against the proofs of this edition.

The passages from Malory used by Mark Twain were checked by three readings of a copy of the Globe edition of Malory against the corresponding sections of the first American edition. The prospectus and the selections in the Century were both read twice against the first American edition. The first English edition was completely read twice against the first American edition; in addition the first English edition, the second English edition, and the Tauchnitz edition were compared in a partial collation covering the preface through chapter 6, chapters 21 through 25, and chapter 41 through the end of the book.

The first American edition was read against a copy of the 1899 edition, and all variants between these editions were checked against a copy of the Harper 1896 edition. The passages marked by the American Publishing Company’s proofreader in the copy of the “Royal Edition” now at Yale were checked by comparing a microfilm of the marked pages in the Yale copy with the same passages in copies of the first American edition, the Harper 1896 edition, and a copy of a Harper impression of the American Publishing Company/Harper 1899 edition.

Machine collations4 of the following editions of A Connecticut Yankee were performed: two copies each of the 1889 and 1891 impressions of the first American edition; one copy of the first Canadian edition (collated against a copy of the first American edition); one copy each of early and late impressions of the 1896 edition; and one copy each of early and late impressions of the 1899 edition.5 Machine collations were also carried out of three [begin page 627] copies of the prospectus and of the passages quoted by Mark Twain and the indexes in copies of five impressions of the Globe edition of Malory.

The following editions were collated at the University of Iowa Textual Center: two copies of the prospectus;6 four copies of the 1889 impression of the first American edition, including one from Special Collections, University of Iowa, and three that are not now identifiable; one unidentified copy of the 1891 impression of the first American edition; and two unidentified copies of the first Canadian edition, one each of the first and second impressions (Toronto: G. M. Rose and Sons, 1889, and Rose Publishing Company, 1890).