In Capgrave's Chronicles of England is a reference to the "londe of Iude," but this is misspelt "Inde" in the edition published in the Master of the Rolls' Series in 1858. Here is a simple misprint caused by the misreading of I for J and n for u; but this can easily be set right. The indexer, however, has enlarged it into a wonderful blunder. Under the letter I is the following curious piece of information:
"India ... conquered by Judas Maccabeus and his brethren, 56"!!
Many more instances of this confusion of the letters u and n might be given, some of them causing permanent confusion of names; but two (which are the complement of each other) will suffice.
George London was a very eminent horticulturist in his day, who at the Revolution was appointed Superintendent of the Royal Gardens; but he can seldom get his name properly spelt because a later horticulturist has made the name of Loudon more familiar. In fact, I was once called to account by a reviewer who supposed I had made a mistake in referring to London instead of Loudon. The reverse mistake was once made by the great Duke of Wellington. C. J. Loudon (who wrote a very bad hand) requested the Duke to let him see the Waterloo beeches at Stratfieldsaye. The letter puzzled Wellington, who knew nothing of the horticulturist, and read C. J. Loudon as C. J. London, and beeches as breeches; so he wrote off to the then Bishop of London (Dr. Blomfield) to say that his Waterloo breeches disappeared long ago.
2. Classification within the alphabet.—Examples have already been given where the arrangement of the book is followed rather than the alphabetical order; but these were instances of bad indexing, and sometimes a good indexer fails in the same way, thus showing how important is good arrangement. An index of great complexity, one full of scientific difficulties, was once made by a very able man. The précis was admirable, and the various subjects were gathered together under their headings with great skill—in fact, it could not well have been more perfect; but it had one flaw which spoiled it. The nature of the index necessitated a large number of subdivisions under the various chief headings; these were arranged on a system clear to the compiler, and probably a logical one to him. But the user of the index had not the clue to this arrangement, and he could not find his way through the complicated maze; it was an unfortunate instance of extreme cleverness. When the index was finished, but before it was published, a simple remedy for the confusion was suggested and carried out. The whole of the subdivisions under each main heading were rearranged in perfect alphabetical order. This was a heroic proceeding, but it was highly successful, and the rearranged index gave satisfaction, and the same system was followed in other indexes that succeeded it.
3. Variety of alphabets.—An index should be one and indivisible, and should not be broken up into several alphabets. Foreigners are greater sinners against this fundamental rule than Englishmen, and they almost invariably separate the author or persons from subjects. Sometimes, however, the division is not very carefully made, for in the Autoren Register to Carus' and Engelmann's Bibliography of Zoology may be found the following entries: Schreiben, Schriften, Zu Humboldt's Cosmos, Zur Fauna. Some English books are much divided. Thus the new edition of Hutchins's Dorset (1874) has at the end eight separate indexes: (1) Places, (2) Pedigrees, (3) Persons, (4) Arms, (5) Blazons, (6) Glossarial, (7) Domesday, (8) Inquisitions.
The index to the original quarto edition of Warton's History of English Poetry (1774) has six alphabets, but a general index compiled by Thomas Fillingham, was published in 1804, uniform with the work in quarto. The general index to the Annual Register has as many as fourteen alphabets. The general index to the Reports of the British Association is split up into six alphabets, following the divisions of each volume.
4. Want of cross references.—Although an alphabetical index should not be classified, yet it is necessary to gather together the synonyms, and place all the references under the best of these headings, with cross references from the others. For instance, Wealth should be under W, Finance under F, and Population under P; and they should not all be grouped under Political Economy, because each of these subjects is distinct and more conveniently found under the separate heading than under a grouped heading. On the other hand, entries relating to Tuberculosis must not be scattered over the index under such headings as Consumption, Decline, and Phthisis, but be gathered together under the heading chosen, with cross references from the others. In bad indexes this rule is invariably broken, and it must be allowed that the proper carrying out of this rule is very difficult, so that where it is invariably adopted, we have one of the best signs of a really good index. Bad indexers are usually much too haphazard in their work to insert cross references.
The careful use of cross references is next in importance to the selection of appropriate headings. Great judgment, however, is required, as the consulters are naturally irritated by being referred backwards and forwards, particularly in a large index. At the same time, if judiciously inserted, such references are a great help. Mr. Poole says, in an article on his own index in the Library Journal: "If every subject shall have cross references to its allies, the work will be mainly a book of cross references rather than an index of subjects." He then adds: "One correspondent gives fifty-eight cross references under Mental Philosophy, and fifty-eight more might be added just as appropriate."
The indexer should be careful that his cross references are real, but he has not always attended to this. In Eadie's Dictionary of the Bible (1850) there is a reference, "Dorcas see Tabitha," but there is no entry under Tabitha at all.