Johnson, Charles 
 
 The Mechanical Processes 
 of the Historian
 
 IRVINE
 
 HELPS FOR STUDENTS OF HISTORY 
 
 No. 50 
 
 The 
 
 Mechanical Processes 
 
 of the Historian 
 
 6/ 
 CHARLES JOHNSON, M.A., FS.A. 
 
 HISTORICAL ASSOCIATION
 
 THE MECHANICAL PROCESSES OF 
 THE HISTORIAN
 
 HELPS FOR STUDENTS OF HISTORY, No. 50 
 
 Edited by C. Johnson, M.A., H. W. V. Temperley, M.A., 
 and J. P. Whitney, B.D., D.C.L. 
 
 THE 
 
 MECHANICAL PROCESSES 
 OF THE HISTORIAN 
 
 BY 
 
 CHARLES JOHNSON, M.A., F.S.A. 
 
 LONDON 
 
 SOCIETY FOR PROMOTING 
 CHRISTIAN KNOWLEDGE 
 
 NEW YORK AND TORONTO I THE MACM1LLAN CO. 
 1922
 
 6. a as 
 MS 
 
 H4
 
 CONTENTS 
 
 Introductory 
 
 PAGE 
 
 I. The Search for Evidence - - - 8 
 
 (1) PRINTED BOOKS - - - - 9 
 
 (2) MANUSCRIPTS AND MONUMENTS - - 10 
 
 II. Formation of Collections - - - 13 
 
 (1) PLAN AND ARRANGEMENT - - - 13 
 
 (2) INDEXING - - - - - 15 
 
 III. Book Production - - - - - 22 
 
 ( 1 ) PREPARATION OF MANUSCRIPT - - - 22 
 
 (2) PROOF-CORRECTING - - - - 25 
 
 IV. A Short List of Books - - - - 28
 
 THE MECHANICAL PROCESSES 
 OF THE HISTORIAN 
 
 INTRODUCTORY 
 
 The mechanical side of historical work is doubtless 
 of less importance than those aspects of it which 
 have been dealt with in other " Helps for Students 
 of History," such as Mr. Crump's " Logic of History " 
 or- Dr. Marshall's " Historical Criticism of a Docu- 
 ment." We may liken it to the bridle of Pegasus, 
 since without it the historian would be at the mercy 
 of his materials. The unhappy " savant " buried 
 alive in an avalanche of multicoloured " fiches," whom 
 M. Anatole France describes in the preface to " L'He 
 des Pingouins," is a parable of the consequences of 
 neglecting this humble side of historiography. 
 
 It is true that the parable may be applied with even 
 greater profit in another way. The mastery of the 
 historian over his evidence, which it enjoins, depends 
 far more on mental than on mechanical capacity. The 
 most superb " technique ' : will not make a little 
 painter into a great one, nor turn the faithful gather- 
 ing, sifting, and classification of evidence into History. 
 Clio, the Muse, must still be invoked. We cannot 
 reduce to a formula that grasp of the essential facts 
 which atones for countless inaccuracies of detail, nor 
 yet the gift of style without which historical writing 
 is sapless and unfruitful. It would be well if someone 
 
 7
 
 MECHANICAL PROCESSES OF THE HISTORIAN 
 
 could write a " Rhetoric of History " which might at 
 least warn the beginner how not to write. But it 
 would be better if more could be done to keep him 
 from trying to amass a larger quantity of material, 
 or to essay a more difficult problem than his mind 
 is capable of handling to advantage. Much labour 
 is wasted on ambitious " theses," which would be 
 better bestowed on the editing of some historical 
 text, or the description of some small class of original 
 documents. 
 
 But however limited the task which the writer sets 
 himself, it will be lightened by a knowledge of the 
 mechanical operations involved. These are, in their 
 order, the search for evidence, its collection and arrange- 
 ment, and the production of the printed book. 
 
 I. THE SEARCH FOR EVIDENCE. 
 
 It is unnecessary to discuss in this place the relative 
 authenticity of the main classes of material: that is 
 a part of the logic of history. For our purpose the 
 distinction to be drawn is that between printed sources 
 which are accessible in many places, and manuscripts 
 or monuments which must be examined where they lie. 
 
 Here the first rule is to exhaust the printed material 
 first, otherwise much time will be spent in doing 
 over again work which has already been done suffi- 
 ciently for the purpose. Here, too, will often be 
 found indispensable guides to the nature and the 
 whereabouts of the unpublished material. 
 
 On the other hand, printed books should not be 
 
 8
 
 THE SEARCH FOR EVIDENCE 
 
 accepted as accurate without criticism or verification. 
 The amount of this varies with the extent of the 
 author's power to exercise it, as well as with the defects 
 of his printed authorities. Thus, in writing a history 
 of the Crusades it may be necessary for him to employ 
 Arabian evidence, without being an Orientalist. In 
 such a case the only course is to take it at second hand 
 from an accredited authority, and say he has done so. 
 But he is presumably able to check the accuracy of 
 his Western sources, and may, on occasion, be forced 
 to examine the original MSS. 
 
 (1) Printed Books. 
 
 The readiest approach to printed material is through 
 Bibliographies, of which there are so many, old and 
 new, that there are even bibliographies of historical 
 bibliography. It is thus, as a rule, easy to find out 
 whether a bibliography exists of the country, period, 
 or subject in question. 
 
 Bibliographies themselves are of two types, select 
 or comprehensive. The Select Bibliography is more 
 useful to the beginner, especially if it gives brief 
 descriptions of the books named, as well as their 
 authors, titles, sizes, and dates of publication. It 
 should be subdivided according to the periods and 
 subjects dealt with, and should have a good index, 
 and cross-references to the main description of each 
 book from the several lists in which it may happen to 
 be mentioned. 
 
 The Comprehensive Bibliography is more useful 
 when some progress has been made with the subject, 
 and the student begins to have some idea of the
 
 MECHANICAL PROCESSES OF THE HISTORIAN 
 
 references which he can safely omit to look up. To 
 this class we must assign the subject-catalogues of 
 great libraries, or of the works published in particular 
 countries. Even more useful are the catalogues of 
 special collections brought together with a definite 
 object, such as the Thomason Tracts, or the three 
 collections in the British Museum relating to the 
 French Revolution. 
 
 The great Literary Histories of various countries 
 have a special value, since they contain valuable 
 criticisms of the works described and often give much 
 information about unprinted material. 
 
 To supplement the bibliographies and bring them 
 up to date, we must have recourse to periodicals. 
 The easiest to use are those which have good indexes 
 covering periods of ten or twenty years. These fill 
 up the gaps between the point where the printed 
 bibliographies stop and that at which the search 
 begins. The reviews which they contain often help 
 by giving a clue to other books earlier than those 
 actually criticised. 
 
 Still more useful, when they exist, are the lists of 
 books appended to modern general histories, and to 
 articles in biographical dictionaries and encyclopaedias. 
 In every case it will be found that one clue leads to 
 another : the finished history leading back to its sources, 
 and these in turn suggesting openings for enquiry as 
 much by what they omit as by what they supply. 
 
 (2) Manuscripts and Monuments. 
 
 The same considerations apply to the search for 
 unpublished material, and many of the books which 
 
 10
 
 THE SEARCH FOR EVIDENCE 
 
 served as a guide to what is published also contain 
 indications of unpublished material. 
 
 There are general guides to the Archives and 
 Museums of the world, as well as to its Libraries, 
 from which can be gathered their size and general 
 character and the conditions on which students are 
 admitted. From these, or from the catalogues of such 
 libraries as the British Museum, indications can be 
 obtained of the guides or printed catalogues to the 
 several collections. These should always be con- 
 sulted, if possible, before time and money is spent on a 
 personal visit. In cases of doubt it is wise to write 
 to the custodians for information before going. 
 
 There is less information about private collections 
 and more difficulty in obtaining access. In England 
 much help can be got from the reports of the Historical 
 MSS. Commission, although they do not deal system- 
 atically with all the collections known to exist, but 
 only with those which have from time to time been 
 brought to the notice of the Commissioners. Such 
 collections, after being reported on, are often dispersed 
 by sale or owing to the death of the owner, and must 
 be traced by means of sale-catalogues or genealogical 
 enquiries. Nor is access always easy, even when the 
 owner is a corporation: for there may be no funds 
 applicable to the payment of someone to supervise 
 the consultation of the documents by students. This 
 should be borne in mind when complaints are made 
 of the excessive fees demanded for access to the MSS- 
 preserved (for example) in Chapter Libraries. It is 
 much to be desired that by co-operation of the Com- 
 mission and of local Archaeological and Record Societies 
 an attempt should be made to obtain a general survey 
 
 11
 
 MECHANICAL PROCESSES OF THE HISTORIAN 
 
 of the material in private hands, and to induce its 
 owners to lend their MSS. to suitable public libraries 
 for consultation by students under conditions more 
 convenient to them and safer for the MSS. than a 
 private library permits. 
 
 A second class of guides is concerned with the 
 materials for the history of particular countries or 
 subjects, and these incidentally provide information 
 as to the existence and whereabouts of material which 
 can be used for the history of other countries and 
 subjects. Some of these will be indicated in an 
 appendix. 
 
 The guiding principle in the search for unpublished 
 material is to have a clear idea of the nature of the 
 evidence sought, and of the reason why it should be 
 in a particular place of deposit. This may involve 
 points of political or institutional history, or of 
 biography or genealogy. 
 
 Thus, for instance, the documents relating to the 
 history of a particular country must be sought among 
 those of the various kingdoms or empires of which it 
 has at various times been a part as well as within its 
 own borders. But it is futile to search in London for 
 records of the local administration of Guienne except 
 in so far as they concerned the central Government. 
 Documents, again, have often been transported from 
 one country to another when they have been necessary 
 for the administration of conquered territory, and 
 other documents kept in the same repository have 
 been transferred with them by mistake. Thus it 
 happens that certain of the records of Flanders are 
 divided, on no clear principle, between Brussels and 
 Lille. But in the main, official documents must be 
 
 12
 
 THE SEARCH FOR EVIDENCE 
 
 sought either at the last resting-place of the institution 
 which produced them, or in the hands of the body 
 which succeeded to its functions. Thus, the records 
 of the old Committee of the Council for Trade and 
 Plantations are divided between the Colonial Office 
 and the Board of Trade. 
 
 Similarly, the papers of a minister of state will be 
 found either among those of his department or among 
 his private papers. In the latter case the succession 
 to his property must be examined. The papers may 
 be with his descendants, or they may have passed by 
 purchase or bequest to a particular library, or they 
 may have been sold with some one of his houses. 
 Even where no sale takes place a collection may be 
 divided, as in the case of the Harley papers, part of 
 which were carried by a marriage into the possession 
 of the Marquis of Bath, while the rest are among the 
 Portland MSS. 
 
 Despite the truth of this principle the chapter of 
 accidents has to be reckoned with. Auction sales, 
 autograph collectors, and museums which pursue the 
 pestilent practice of picking out the plums of a col- 
 lection, are the worst enemies of the searcher for 
 material. So the net must be spread wide that good 
 luck may supplement good management. 
 
 II. THE FORMATION OF COLLECTIONS. 
 (1) Plan and Arrangement. 
 
 In history, as in the more exact sciences, the first 
 step is to commit the information obtained to writing, 
 and to prepare a means by which it can be readily 
 
 13
 
 MECHANICAL PROCESSES OF THE HISTORIAN 
 
 found: the " Experientia litterata " and the " Tabulae " 
 of Bacon.* Human memory is unequal to the task. 
 We must therefore make MS. collections, and index 
 them in such a way as to be able to lay hands at any 
 time on the evidence of any particular point. 
 
 When these collections consist of extracts from 
 printed or MS. authorities, or of descriptions of monu- 
 ments, or of disquisitions on points of special im- 
 portance, it is well to have them in a form which 
 permits them to be sent to the printer without re- 
 copying, and the rules as to size of paper, margin, 
 and so forth, to be stated later, apply equally in this 
 case. If this plan is followed, the sheets can be filed 
 in any desired order, and put away in filing-covers, 
 tied up with tape into flat bundles, or put in cardboard 
 boxes properly labelled for reference. Smaller divi- 
 sions can be indicated by slips of paper exceeding the 
 normal length, or by the detachable " tabs " which 
 most stationers can supply. This is, on the whole, 
 more convenient than using notebooks, which must 
 either be torn up or recopied. 
 
 References to be looked up or checked can be noted 
 in a small memorandum book, if not too numerous. 
 Another plan is to note them on small slips of paper, 
 a sheaf of which can be conveniently carried about, 
 each slip being torn up when done with, or put away 
 in its proper place if wanted again. 
 
 The titles and names of books and documents used 
 should be most carefully kept on cards or in a book, 
 and the full details should always be noted, in case 
 it should be necessary to print a reference or to include 
 the title in a bibliography. It is particularly im 
 
 * " Novum Organum," Book 1, Aphs. 101 and 102. 
 
 14
 
 FORMATION OF COLLECTIONS 
 
 portant in quoting documents to distinguish between 
 those quoted from a printed or MS. copy and those 
 of which you have examined the original. In all cases 
 a sufficient indication should be given to enable another 
 person to verify your reference for himself. 
 
 Dates should be checked at the earliest possible 
 stage by the use of perpetual calendars and tables of 
 eclipses, councils, regnal years, etc. 
 
 Materials may be arranged either chronologically, 
 or by subject, or by the sources from which they are 
 derived, and if the collection is large all these methods 
 may have to be used. The exact plan must be fixed 
 by the use which is to be made of them. It is wise 
 to make quite certain what the plan is, and even to 
 put it down on paper. Cross-divisions cannot be 
 avoided altogether, and cross-references should there- 
 fore be noted in the proper places lest vital facts 
 should be overlooked. 
 
 (2) Indexing. 
 
 In a large collection, systematic arrangement and 
 occasional cross-references are not of themselves 
 sufficient for control of the material in all its detail, 
 and a certain amount of indexing is required. The 
 collector, however, knows better than a stranger how 
 to find his way about the collection, and can thus save 
 himself some of the labour which the indexer of a 
 printed book must take. But the same principles 
 apply in both cases, and are therefore stated here. 
 The application of them differs very widely, even in 
 books. Thus, a printed collection, like Rymer's 
 " Fcedera," requires the indexing of every personal 
 or local name: in a simple narrative the number of 
 
 15
 
 MECHANICAL PROCESSES OF THE HISTORIAN 
 
 references to these may be smaller, but more care 
 must be taken to index the events and subjects 
 treated. The guiding purpose is to enable the reader 
 to find whatever information he may be in search of. 
 Assuming, then, that a printed book is being indexed : 
 
 (1) It is best to have only one index, of Persons, 
 Places, and Subjects. It is usually easier to consult 
 a single index than three separate indexes. But some 
 classes of entries can, with some saving of labour, be 
 grouped in sub-indexes, if it is made clear that this 
 has been done. For example, in a work on art, a 
 separate sub-index of names of pictures or artists 
 may be more convenient to the reader than the in- 
 clusion of these names in the general index. 
 
 (2) The index should be strictly lexicographical in 
 the order of the catchwords. When many entries 
 occur under the same catchword they should be sub- 
 divided into groups, so that not more than about 
 a dozen references need be looked up to find the passage 
 sought. The main exception to this rule is the case 
 in which there is no means of differentiating the 
 references—fi.gr., " Carlisle, Letters Patent dated at." 
 In such a case, if the searcher has no other guide, 
 there is no means of saving him labour. 
 
 (3) When the same catchword is used in more than 
 one sense, it should be repeated each time the sense 
 changes, and the senses should follow each other in a 
 fixed order. Thus, places should come before persons, 
 and persons before subjects. For instance, an entry 
 under Lincoln might run as follows : 
 
 Lincoln, County of, 24, 66, 105. 
 
 Sheriff of, 27. 
 Lincoln (co. Lincoln), 45. 
 
 Burgesses of, 753. 
 16
 
 FORMATION OF COLLECTIONS 
 
 Lincoln (Maine, U.S.), 15. 
 Lincoln, Bishop of. See Alnwick. 
 Lincoln, Earl of. See Clinton. 
 Lincoln, James, citizen of London, 34, 67. 
 
 John, 53. 
 " Lincoln Green," 578. 
 
 (4) All the references to the same person, place, or 
 subject should be indexed under the same catch- 
 word, with cross-references from the other forms in 
 which they occur. The form to be chosen for the 
 main entry should be, as a rule, that most likely to 
 be familiar to the searcher. Where this is uncertain 
 a definite rule should be adopted, exceptions to which 
 should be sparingly allowed. 
 
 (5) Where the same reference constantly recurs, 
 it is useless to record all the instances. It should 
 be noted as " Passim,'''' or " pp. 1 to 180, passim" 
 as the case may demand. 
 
 (6) Places should be indexed under the form used 
 in the Ordnance Map, or in a standard Gazetteer. A 
 list of standard spellings is being published by the 
 Royal Geographical Society, and should be followed 
 when applicable. Where the identification is un- 
 certain, the form in the text should be followed, with 
 a cross-reference from the standard form of the name 
 supposed to be meant. Where many forms are found 
 in the text and identification is uncertain, the central 
 form should be chosen on the usual principles of 
 textual criticism. Place-names with descriptive pre- 
 fixes such as "Great," "Little," "Long," or 
 " Market " should be indexed under the name without 
 cross-references from the prefixes. 
 
 (7) Persons should be indexed under their family 
 names, if known, with cross-references from their 
 
 17
 
 MECHANICAL PROCESSES OF THE HISTORIAN 
 
 titles or descriptions. For periods in which family 
 names have not begun to be used, it is best to index 
 under the Christian name, avoiding such catchwords 
 as " Alius " or " Fitz," except in such cases as " Fitz- 
 gerald " or " Fitzalan," in which the patronymic has 
 become a family name. 
 
 In such cases as " William, son of Thomas," or 
 : ' John, son of William," it is better to index both 
 Christian names if they occur at a date at which 
 surnames are unusual. If they are rare cases occurring 
 among a multitude of surnames it is enough to treat 
 them as though they had been " Williamson ' : or 
 " Thomson," and index them under " William, son 
 of " and " Thomas, son of " respectively. 
 
 (8) Saints, Popes, Kings, Jews, and Welshmen may 
 be regarded as having no surnames. When they have 
 known surnames, cross-references from these should 
 be given. 
 
 (9) Special difficulty is caused by such descriptive 
 
 names as " John, son of William of London," " Giles, 
 
 prestesservant of Holt," and trade names such as 
 
 " baker " or " smith." Here, other references may 
 
 supply the surname under which the entry should be 
 
 made. If not, both Christian and place-name should 
 
 be indexed — e.g., 
 
 William, John son of, of London. 
 London, John son of William of. 
 Giles, priest's servant of Holt. 
 Holt, Giles prestesservant of. 
 William the smith. 
 Smith, William, etc. 
 
 (10) Where a person is mentioned simply as the 
 holder of an office without reference to his individual 
 character, the entry should be made under the office 
 
 18
 
 FORMATION OF COLLECTIONS 
 
 only. Thus, a will proved before the Archbishop of 
 Canterbury will be indexed as " Canterbury, arch- 
 bishop of, Will proved before," and not under " Islip," 
 supposing him to be the holder of the office and not 
 to be specially named. 
 
 (11) Subjects should not necessarily be indexed 
 under the catchword suggested by the text. These 
 are often not the first which would occur to a searcher. 
 For example, it would be useless to index the " Ems 
 telegram " under " telegram," while it might properly 
 appear under " Franco-German War." The indexer 
 must try to grasp the significance of the statements 
 in the text, and to group them under general heads. A 
 particular case, of which the details are known, can 
 usually be found by a person- or place-reference. Allied 
 subjects must be brought together by cross-references. 
 
 (12) Cross-references are of two kinds, simple 
 or double. The former (See . . .) leads from a 
 synonym or a subdivision to the main subject; the 
 latter (See also . . .) links one subject to another 
 which is distinct but allied. 
 
 (13) Cross-references, whether simple or double, 
 should be used sparingly. For instance, it is not 
 necessary to record every variant in spelling of a 
 person- or place-name. Where two of these come 
 next to each other in alphabetical arrangement, one 
 can be omitted or the two combined in a single cross- 
 reference. In the same way double cross-references 
 of subjects should be limited to the cases which cannot 
 be grouped as subdivisions of a main head, though 
 closely connected with it. 
 
 The mechanical labour of indexing can be lightened 
 by delegating some of the work to an intelligent 
 
 19
 
 MECHANICAL PROCESSES OF THE HISTORIAN 
 
 helper. The system to be followed is essentially the 
 same in all cases, though the methods differ in detail 
 with the size of the index and the tastes of the indexer. 
 
 The index should be begun as soon as the paging 
 is settled, each sheet of a book being indexed before 
 it is " passed for press." This permits corrections to 
 be made where the index, as it often will, reveals 
 mistakes or oversights. 
 
 The person responsible for the index should either 
 prepare the MS. himself or mark on the printed sheet 
 the words to which references are required. A general 
 instruction can be given as to person- and place-names, 
 but subjects must, at all events, be marked, and in 
 cases where names occur in forms differing from those 
 to be used as catchwords, the identifications must be 
 written in the margin. In the same way, subjects, 
 the catchwords for which do not appear in the text, 
 must have them written in. 
 
 The entries must be made on slips, and the slips 
 kept sorted, unless the index is so small as to be easily 
 made in an alphabet-book, which it seldom is. If a 
 typewriter is to be used, the entries should be made 
 on long strips of paper marked at even distances, so 
 that they can be cut up into slips of equal size. It is 
 best to have these strips narrow, so that when they 
 are cut up the entry is across the end of the 
 finished slip. Such slips when made into bundles can 
 more readily be examined than is possible if the entries 
 run the long way of the paper. 
 
 Each word should be marked off as it is entered, 
 and as each page is finished the sheets or slips should 
 be glanced through to make sure that the page numbers 
 are all the same, and that none have been omitted. 
 
 20
 
 FORMATION OF COLLECTIONS 
 
 The slips should be sorted at frequent intervals, 
 duplicate entries being transferred to the slips on 
 which the first references were posted. Some people 
 keep these waste slips for future use, in case a second 
 edition of the book should be needed. 
 
 When the book is all indexed, the slips should be 
 examined to see that they are in order. It is very 
 easy to misplace a slip and thus to get duplicate 
 series of entries. At the same time the larger heads 
 can be gone through and repeats struck out in the 
 subheads. The printer will usually omit them, but 
 it is better not to leave it to him. 
 
 While the index is being made, the slips can be 
 kept in order by having a box or trough with guide 
 cards, like a card -index cabinet. Or they may simply 
 be made up into medium-sized bundles fastened 
 together by indiarubber bands, and these may be 
 packed in order in a box of any kind. 
 
 The actual sorting may be simplified by having five 
 boxes, side by side, each divided into five sections. 
 This provides a compartment for every letter of the 
 alphabet but Z, which is rare enough to be left out- 
 side. The hand soon gets an instinctive feeling of 
 where to place each slip. Needless to say, the groups 
 thus mechanically sorted must be checked through 
 as they are put away, but the mistakes found will be 
 few after the first. 
 
 Before sending slips to the printer it is well to 
 number them with a numbering machine in order to 
 avoid loss or confusion at the printer's. 
 
 21
 
 MECHANICAL PROCESSES OF THE HISTORIAN 
 
 III. BOOK PRODUCTION. 
 
 (1) Preparation of Manuscript. 
 
 Manuscript for the printer should be on one side 
 of the paper only, so that it can be cut into sections 
 for division between the compositors. The paper 
 should be of the same size for convenience of handling 
 and packing, and should be ruled to make it easier to 
 calculate the number of words. It should not be 
 larger than quarto, whether written or typed, since 
 larger sheets are apt to curl up and will not stand 
 upright in front of the compositor. 
 
 A margin of an inch should be left for corrections 
 and headings. 
 
 MS. should be typed, or legibly, but not too legibly, 
 written. A strikingly clear hand is a direct tempta- 
 tion to careless type-setting : and more than one author 
 has found that his worst written MSS. were the best 
 set up. But it is not right to set the printer too hard 
 a task. If a page is much interlineated or corrected 
 it ought to be rewritten. 
 
 If the MS. is written, footnotes can be put in at the 
 bottom of the page without inconvenience. If the 
 author types his own MS., he will possibly find it wise 
 to copy the printer's device of letting the footnotes 
 follow the passages to which they refer, indenting them 
 a little to distinguish them from the text. 
 
 It is well to know how much space any part of a 
 book will occupy in type. The printer will usually 
 give a very accurate estimate or " cast," but a fair 
 guess may be made, if the copy is typed, by counting 
 the number of words to a line and of lines to a page, 
 multiplying them together, and multiplying the result 
 
 22
 
 BOOK PRODUCTION 
 
 by the number of pages. If the MS. is written and 
 the paper not ruled, it is a good plan to count one 
 page in every ten for lines and one line taken at 
 random in each of the pages, and thus obtain averages 
 for the number of words in a page. The approximate 
 number of words should always be written on the 
 outside sheet. 
 
 When the number of words is known it is easy to 
 estimate the size of the book by taking a volume 
 printed in the style preferred and finding out in the 
 same way the average number of words to a page. 
 The figures given in " Whitaker's Almanack " will 
 show how much space will be gained or lost by using 
 larger or smaller type. It should be remembered that 
 blank spaces cost as much as letterpress. 
 
 It is a good plan to leave space in the MS. for any 
 blocks which are to be printed with the letterpress. 
 Blocks are charged for by the square inch, with a 
 certain minimum. If the blocks have already been 
 made, impressions of them can be taken in endorsing 
 ink and pasted on the MS. in their proper places. 
 
 A good deal of trouble may be saved in writing by 
 using the abbreviations commonly recognised in the 
 printing trade: the compositor will in most cases have 
 no trouble in interpreting them, or any special abbrevia- 
 tions of which he has due notice. 
 
 Certain points, such as Title Page, Chapter Headings 
 and Headlines, Index, and Bibliography, cannot be 
 completely settled in MS., but thought should be 
 given to them, and as much as possible done before 
 sending in the MS. For instance, the author should 
 know how he intends to be described on the title- 
 page, and should allow for the index in estimating the 
 
 23
 
 MECHANICAL PROCESSES OF THE HISTORIAN 
 
 size of the book. This will reduce the number of 
 unpleasant surprises awaiting him at a later stage. 
 
 The sheets of the MS. should be numbered, pierced 
 at one corner with a bodkin, and firmly fastened 
 together with string or tape. If they escape from 
 their envelope in the post, they will then be all found 
 together and in order. They should be packed flat 
 if possible; if not, folded once only, the long way of 
 the paper, never rolled. A roll of papers not fastened 
 together at the head is the despair of the publisher's 
 reader, and he will certainly be correspondingly slow 
 to read it, and prejudiced against its contents. 
 
 Manuscript, or proof which has taken a long time 
 to correct, should always be sent by registered post. 
 This is not necessary when there are few corrections, 
 since the printer has a duplicate of the uncorrected 
 proof. 
 
 A word must be added on the subject of Appendices. 
 They are best avoided, but it often happens that a 
 question arises which cannot be discussed in the text, 
 or within the limits of a moderate footnote. In such 
 a case a reference should be given in the body of the 
 work to the appendix, and the appendix provided 
 with a reference to the page of the text to which it 
 relates. A special case is the Bibliography, which it 
 is now usual to add to every serious work. This should 
 be cut as short as possible, and limited to works 
 frequently quoted. It should always mention the 
 place and date of publication of the edition used, but 
 the titles should be cut as short as can be without 
 ambiguity. A mere parade of authorities is as wasteful 
 as it is annoying to the reader. 
 
 24
 
 BOOK PRODUCTION 
 
 (2) Proof-Correcting. 
 
 The MS. sent to the printer will usually return in 
 what is called " slip- " or " gahVy -proof," arranged, 
 that is, in columns but not in pages. All possible 
 corrections should be made at this stage; since, after 
 the slip is made up into pages and the " formes " 
 locked up, it takes much more time to make any 
 alteration. Indeed, the insertion of a line or two of 
 additional matter may upset the paging of a whole 
 sheet, or even of two or three. All corrections, not 
 purely due to the compositor, are paid for by time, 
 and add to the cost of production of the book. As 
 the printer's reader is not always to be depended upon 
 to see that the proof corresponds with the copy, it is 
 best to distinguish " author's corrections ' : from 
 ' printer's errors " by using a different coloured ink 
 for them. If the publisher afterwards makes an extra 
 charge on the ground that the " author's corrections ' : 
 exceed the usual allowance, this plan makes it easy 
 to see what was their actual extent. 
 
 The accompanying example shows the usual marks 
 made on the proof to indicate corrections. Be very 
 careful that they are legible and not ambiguous. Note 
 also that words in small capitals which are to have 
 larger initial letters can be indicated by marking the 
 initials with three lines as for capitals and the rest 
 with two lines as for small capitals. 
 
 Remarks addressed to the printer and not intended 
 to be printed should be enclosed in a ring. Insertions, 
 indicated by a connecting line, should have the loop 
 enclosing them left open at one end. 
 
 If you are in any doubt whether or not a correction 
 
 25
 
 ~tu> 
 
 MECHANICAL PROCESSES OF THE HISTORIAN 
 
 Page showing corrections. 
 
 £J /JDo not try to correct the faults of hurried making-ready/by 
 
 A a weak impression, and by carrying an excess of ink to (nide 
 
 @ A the weakness/ Excess of ink fouls the rollers, clogs the 
 
 type, and makes the printed work smear or set off. A good irft-c 
 print cannot be had w hen the impression is so weak that the 
 paper (j;ouchesp jarelylthe ink on the types and is not pressed 
 against the types. There must be force enough to transfer y. 
 the ink not only on to the paper^but into the paper. A firm ,/ tayr^ 
 tAzjf" £>/ impression sLpuJt i be had, even if the paper be indented. / 
 
 ' The amount of impression required will largely depend on 
 (q) the making-ready. MJth carefufciaking- ready, ^repression s^ L- 
 
 * may be light; roughly"and hurriedly done, it must be hard/ \l 
 -c,C .Indentation IS evidence of wear of type. ' The spring and i y 
 
 qJ resulting friction of an el/stic impression surface as most felt 
 - , / where there is least resistance — at the upper and lower ends 
 7/ of lines of /ype, where they begin to round off. It follows 
 
 that the laving of time that may be gained by hurried and tit 
 rough making-ready must be offset by an increased wear of ' 
 type. '[That impression isJpesc for preventing wear of type nve.1 (5) 
 * which,i8 confined to its surface and never laps over mstm its '37 
 
 edges. But this perfect surface impression is possible only A? 
 , . on a large forme with new type, sound, se& packing, and -AafaC 
 J // *-/ ample time for^Anakingjready.^ If types are worn, the in- ^J. / 
 ' // / dentation of the paperby impression cannot be entirely 
 a (/^ prevented/ ^ood presswork docs not depend entirely upon *j 
 ,, ' the/pressAr machine, neither on the work man, nor on the ' 
 
 kvAt/2 
 
 vuty presswr macuiim, iieiiutjr uii me.ffuri nuaii, nur on 1,11c n_^, 
 
 materials. Nor will superiority in any ««tfpoint compensate * c/y 
 
 j • for deficiency in another : new type will suffer from a poor ' 
 
 OlJ ./roller, and careful making-rea^y is thrown a way if poor ink 'Zr /Ur; A t a 
 
 . // be used/ It is necessary that all the mat eria ls shall be ? 
 
 ?y ' good, that t/fey should be adapted to each *6ther and fitly , 
 
 / / used. A good workman can do much with poor"materials, L 
 Iffy but a negle'ct to comply with one condition ofteTTproduces 
 / as bad a result as the neglect of alO 
 
 £447? tPt Of the foregoing facts are carefully studied mafiy diffkul- fU, 
 ties will be overcome in obtaining really good work 
 
 trs.= transpose. stet, with dots under word, cancels correction. 
 
 l.r.= lower case — i.e., small letter for capital. 
 
 The curly mark indicates a type upsido-do.vn. 
 
 $=,!,/,. I = a space. A = cant, marks omissions. 
 
 Stops (points and colons) are ringed. 
 
 26
 
 BOOK PRODUCTION 
 
 is intelligible, write the word in the margin as you 
 
 intend it to be printed, mark it " To read ," and 
 
 put a ring round the whole. 
 
 It is a good plan to have a specimen page set up 
 before finally sending the book to press. General 
 instructions can then be given as to spelling, capitals, 
 and style of printing. The printer will usually be 
 competent to carry these out, and much correction in 
 slip will be saved. 
 
 If the corrections are very numerous, or much new 
 matter is inserted, mark the proof " Revise," with the 
 date and your initials, and send it back to the printer. 
 If it is nearly right, go over it carefully to see that the 
 Title-page, Contents, and Chapter-headings are pro- 
 vided, and arrange the Headlines. The most satis- 
 factory plan for the latter is to have the title of the 
 book as the headline of the left-hand page, and the 
 title of the chapter or section, or an indication of the 
 contents of the two pages on the right-hand. 
 
 If 3'ou wish to see the proof again, mark it " Revise 
 in page." No important corrections should be made 
 after this, particularly none that involve ' over- 
 running " — i.e., alteration of the page-divisions. If 
 any matter has to be cut out, new matter of corre- 
 sponding length must be inserted on the same page. 
 When done with, the proof, whether in slip or page, 
 should be marked " Press," with initials and date, and 
 returned to the printer. 
 
 It is usual to send proofs in duplicate or triplicate, 
 and it is almost always possible to have extra proofs 
 sent to friends, who may have undertaken to read 
 them and assist in correcting. In such cases there 
 is usually one " marked proof," which has been care- 
 
 27
 
 MECHANICAL PROCESSES OF THE HISTORIAN 
 
 fully read and corrected by the printer's reader. It 
 may also contain queries from the publisher. This 
 should always be studied with attention and returned 
 with corrections. If you wish to keep a note of the 
 corrections you have made you should copy them on 
 one of the duplicate proofs. The final proof, with 
 your, and the publisher's, instruction " Press," is kept 
 by the printer as his voucher in case any question 
 should arise as to payment for corrections. 
 
 The queries on the " marked proof " may be either 
 typographical — for instance, when a spelling is doubt- 
 ful — or material. In a first-class printing and publish- 
 ing business the proofs are often read by persons fully 
 competent to understand and criticise the books they 
 print. So you should not be surprised to find a query 
 as to your meaning when you state a fact inconsistent 
 with the usual informed opinion on the subject. But 
 generally a query means that your words can be 
 understood in more than one way, and that the printer 
 wishes you to remove the ambiguity. On no account 
 should such queries be left unanswered. 
 
 IV. A SHORT LIST OF BOOKS. 
 
 1. Printed Sources. 
 
 It is intended to include in this series a short guide 
 to the bibliography of History, but it may not be 
 amiss to give here the titles of a few books which the 
 writer has found useful. 
 
 28
 
 A SHORT LIST OF BOOKS 
 
 General Bibliographies. 
 
 Langlois (Charles V.): Manuel de Bibliographie 
 historique. (1) Instruments bibliographiques. 
 Paris, 1896. 
 
 Herre (Paul): Quellenkunde zur Weltgeschichte. 
 Leipsic, 1910. 
 
 A very unequal book, but useful as attempting to 
 cover the whole field. 
 
 Chevalier (Cyr Ulysse Joseph): Repertoire des 
 Sources historiques du Moyen-Age. (i.) Bio- 
 bibliographie ; (ii.) Topo-bibliographie. Paris and 
 Montbeliard, 1883-1907. 
 
 A vast and uncritical assemblage of references. 
 
 Potthast (August): Bibliotheca historica medii sevi. 
 2 vols. Berlin, 1896. 
 
 A guide to the great printed collections — alphabetical 
 index of chronicles, etc., with particulars of publica- 
 tion — index of lives of saints. Indispensable. 
 
 Wolf (Gustav): Einfiihrung in das Studium der 
 neueren Geschichte. Berlin, 1910. 
 
 Very good on the difference in kind between mediaeval 
 and modern sources. 
 
 Much help may also be got from periodicals, which 
 bring the ordinary reference books up to date — e.g. : 
 
 Historical Association : Annual Bulletin of Historical 
 Literature. London, 1911-1922. 
 
 This Society also issues to its members useful biblio- 
 graphies of particular subjects, mainly from a teacher's 
 point of view. 
 
 English Historical Review. London, 1885-1922. 
 
 Jahresberichte der Geschichtswissenschaft. 
 Berlin, 1878-1922. 
 
 Revue Historique. Paris, 1875-1922. 
 
 History. Quarterly Journal of the Historical Associ- 
 ation. London, 1916-1922. 
 
 29
 
 MECHANICAL PROCESSES OF THE HISTORIAN 
 
 American Historical Review. New York, 1896-1922. 
 Notes and Queries. London, 1850-1922. 
 
 This is, of course, a mere lucky bag, but is often 
 useful. 
 
 The bibliographies of the larger histories, such as 
 the Cambridge Mediaeval and Modern Histories, should 
 also be consulted. 
 
 Particular Countries and Subjects. 
 
 Gross (Charles) : The Sources and Literature of English 
 History to about 1485. 2nd ed. London, 1915. 
 
 Monod (Gabriel): Bibliographie de l'Histoire de 
 France; Catalogue jusqu'en 1789. Paris, 1888. 
 
 Molinier (Auguste): Les Sources de l'Histoire de 
 France jusqu'en 1815. Paris, 1901. 
 
 Pirenne (Henri): Bibliographie de l'Histoire de 
 Belgique jusqu'en 1830. 2nd ed. Brussels, 1902. 
 
 Dahlmann (Friedrich Christoph) and Waitz (Georg): 
 Quellenkunde der deutschen Geschichte. 8th ed. 
 Leipsic, 1912. 
 
 Local Societies. 
 
 Many valuable articles are hidden in the proceedings 
 of local societies. The following guides are useful for 
 the periods which they cover : 
 
 Gomme (Sir George Laurence) : Index of Archaeological 
 Papers, 1665-1890. London, 1907. 
 
 Index (Annual) of Archaeological Papers, 1891-1908. 
 London, 1892-1912. 
 
 Lasteyrie (Robert de): Bibliographie generate des 
 Travaux historiques et archeologiques publiees par 
 les Societes savantes de la France, 1886-1904. 
 5 vols. Paris, 1898-1911. 
 
 30
 
 A SHORT LIST OF BOOKS 
 
 General Bibliographies and Library Catalogues. 
 
 Sonnenschein (William Swan): The Best Books. 
 2nd ed. 1903. 
 
 A revised edition is in progress. 
 
 Fortescue (George Knottesford) : Subject Index of 
 
 Modern Works added to the British Museum. 
 
 (Quinquennial volumes), 1880-1920. 10 vols. 
 London, 1886-1922. 
 
 Wright (Charles Theodore Hagberg): Subject Index 
 of the London Library. London, 1909. 
 
 Much help can also be got from the " Dictionary of 
 
 National Biography," and from the various historical 
 
 and biographical dictionaries, as well as from the 
 
 ' Encyclopaedia Britannica " and the German " Con- 
 
 versations-Lexika. " 
 
 2. Unprinted Sources. 
 
 Among the " Helps for Students of History " will 
 be found several dealing with different branches of 
 historical work as well as with the contents of libraries 
 and archives. For these it is enough to refer to the 
 advertisements on the covers of this pamphlet. We 
 may add the article on " Records " in the " Ency- 
 clopaedia Britannica " and the information contained 
 in the annual volumes of " Minerva." See also — 
 
 Scargill-Bird (S. R.): Guide to the Principal Classes 
 of Documents in the Public Record Office. 3rd 
 ed. London, 1908. 
 
 A new edition is in preparation. 
 
 Livingstone (M.): A Guide to the Public Records of 
 Scotland. London, 1905. 
 
 Wood (Herbert) : Guide to the Records deposited in 
 the Public Record Office of Ireland. Dublin, 
 1919. 
 
 31
 
 A SHORT LIST OF BOOKS 
 
 Langlois (Charles V.) and Stein (Henri) : Les Archives 
 de l'Histoire de France. Paris, 1891. 
 
 Stein (Henri): Bibliographie generale des Cartulaires 
 francais ou relatifs a l'Histoire de France. Paris, 
 1907. 
 
 and the various guides to the sources of American 
 history issued by the Carnegie Institution. 
 
 Chronology. 
 
 A set of " English Time Books " will form part of 
 this series. Others are: 
 
 Grotefend (Hermann): Taschenbuch der Zeitrech- 
 nung des deutschen Mittelalters und der Neuzeit. 
 Hanover, 1898. 
 
 Cappelli (A.): Cronologia e calendario perpetuo. 
 Milan, 1906. 
 
 Bond (J. J.): Handy Book for verifying Dates. 4th 
 ed. London, 1889. 
 
 and the larger works of Ideler, Grotefend, Mas-Latrie 
 and the " Art de Verifier les Dates." 
 
 3. Book-Production and Indexing. 
 
 Brown (George E.): Indexing. A Handbook of 
 Instruction. London, 1921. 
 
 Jacobi (Charles T.) : Some Notes on Books and Printing. 
 3rd ed. London, 1903. 
 
 Collins (F. Howard) : Authors' and Printers' Dic- 
 tionary. 5th ed. London, 1921. 
 
 PRINTED IN' GREAT BRITAIB BY 
 
 HILLING AND SONS, LTD., GUILDFORD AND BBHBH
 
 | DATE DUE 
 
 RI€D NOV 
 
 2 7 1978 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 CAVLORD 
 
 
 
 PRINTED IN U 9 A 1
 
 UC SOUTHERN REGIONAL LIBRARY FACILITY 
 
 " I !! II! Ill II II lllll I I 1 1 III 
 
 ii i; iiiii 1 1 1 1 mi 
 
 AA 000 634 930 2