Page:The War with Mexico, Vol 1.djvu/435

406

determines when, where, and with what forces a battle shall be fought, and tactics directs the battle itself, It follows that on the defensive one should protect one's communications, force the enemy to send out detachments and rout or annoy these; and, if such methods have no saving effect, retire to a position strong by nature, art, or both, and, while defending it, watch for opportunities to strike. One may consult on those points and on the text to which this note refers Clausewitz, On War, book i, chap. 7; Jomini, Précis, i, 470-2; Murray, Reality of War, 21, 77; Donat, Strategical Science, 275 — 6; Henderson, Science of War, 19720; Griepenkerl, Applied Tactics 116; Goltz, Conduct of War, 1042; ''Jour. Milit. Serv. Institut, 1908, p. 31; Am. Hist. Rev,'' xi, 3884

5. As the author was compelled to depart in many cases from the familiar method of referring to the sources, he feels bound to explain how these were handled. All the material, condensed as much as it safely could be, was marked in the margin with Roman figures, indicating to what chapter each sentence or larger section would belong. Then the sections were copied into packets, each oi which contained all the material of a chapter. Next the material of each packet was analyzed into topical items, and the items were numbered with Arabic figures. In Writing a chapter the author placed after each sentence (or, if the case demanded, after each clause, phrase or word) the Arabic figures numbering the items upon which it rested. These figures were retained through the successive revisions until the MS. was ready to print, and were used in the re-examination of the work. By this routine every document was considered at least five times. Of course care was taken at all stages to ensure correct copying; yet in the final revision the author went back, unless there was a good reason for not doing so, to originals or to trustworthy copies from the originals — doing this not merely to verify the references but also to see, in the light of the completed investigation of the subject, whether he had omitted or misunderstood anything of importance in making notes and condensations. The text and remarks as written looked thus:

 "Gen Patterson once asserted that Volunteers were no worse than regular 578, but the evidence of other officers 83 — 4 340 385 564 568 1266 and of the Mexicans 84 87 340 489 582 1236 was overwhelmingly against him. Regulars committed offences, but these appear to have been commonly mere pilfering 181 562, and to have been chargeable mostly to fresh recruits 83 1103 It is probable, however, that the Volunteer troops often bore the blame 109 for acts done by soldiers dishonorably discharged 659 9.31, deserters 582 601 659 639 945, teamsters and other civilian employees 358 1315 and the many "black legs" 377 and "human vultures" 1292 who followed the army 365 568 639. The. great difficulty was to identify the culprits 1315, Mexicans were often afraid to testify against our soldiers 1266."

The saving of space and labor that resulted from the grouping of citations can easily be illustrated. Take the case of ten statements resting upon ten documents, which may be represented by the first ten letters of the alphabet Were the usual method followed the references might be: acj, dfghi, be, adeghj, fij, bcf, adeg, cfh, i, bdghj; Whereas by the group method the references would be abcdefghij — in the former case 35, in the latter 10. The higher cost of the former method might have enforced the omission of all references from the printed volumes, if indeed the increased labor of writing, copying, and proof-reading the references had not