Page:Englishhistorica36londuoft.djvu/178

 170 THE GENEALOGY OF THE April to have known as kings Creoda and Cwichelm, who are omitted from the regnal table. Creoda's name, indeed, is actually inserted in some of our manuscripts both in the preface and in the annal of 855 ; while Cwichelmeshlaew, in Berkshire, has preserved the name of Cwichelm, who, as he had a heathen burial, must be distinguished from Cwichelm Cynegilsing, a con- vert to Christianity. Thus the annal of 593, ' Here Ceawlin and Cwichelm and Crida perished ', may be due to a harmonistic effort of the chronicler's. There is one characteristic of these early annals which deserves notice. The names of the kings and princes whose battles are recorded usually occur in pairs. Sometimes the names of the pairs are alliterative. The practice, already x alluded to, of arrang- ing the generations in alliterative pairs continued, indeed, until the time of King Edmund, 2 but its original purpose was, doubtless, mnemonic, and accordingly it belonged to a time when historical records were entrusted, not to writing, but to the human memory, for their preservation. Such records, doubt- less, took the form of alliterative verse. The battles which belong to the alliterative class are : 495. Cerdicesora 508. Natanleah L v, -. MA ~ ,. ,, fCerdic and Cynric. 519. Cerdicesford J 527. Cerdicesleah 556. Beranburh Cynric and Ceawlin. 568. Wibbandun 577. Deorham I Ceawlin and Cuthwine, or Cutha. 584. Fethanleah J 614. Beandun. ^ ^ Cwichelm 628. Cirenceaster) J & Mr. Chadwick thinks that the traditional number of years for the reign of Ceawlin was seventeen. 3 If so, its extension to thirty-two may be due to the octennial arrangements of his battles. The battles of Wibbandun, Deorham, and Fethanleah divide Ceawlin's reign into four periods of approximately eight years each, the fourth octennium being completed by Woddes- beorh (592), when Ceawlin was ' driven out '. The feat of Cuth- wulf, recorded in the annal of 571, by which he dispossessed the Britons of Lenbury, Aylesbury, Bensington, and Eynsham, does not belong to the octennial arrangement. It is hardly credible that Buckinghamshire was still in the hands of the Britons in 1 p. 164, n. 2, above. 2 Perhaps even to the end of the dynasty. Edmund and Edgar are not strictly alliterative, since the first syllables are identical, but a similar identity occurs much earlier in the case of Cenferth and Cenfus. 3 Op. cit. p. 24 n.