The Peoples of Middle-earth (33 page)

And Atanamir lived to a great age, clinging to his life beyond the end of all joy; and he was the first of the Numenoreans to do this, refusing to depart until he was witless and unmanned, and denying to his son the kingship at the height of his days.

The much greater age of Atanamir must imply that all the other kings died by act of their own will long before the end of their physical span, and thus allowed their sons a period of rule equivalent to their own. It would be mistaken to press this early and experimental text too closely on the matter, but it certainly suggests a difference from the developed conception in The Line of Elros, where it is said (Unfinished Tales p. 218) that it was the custom 'until the days of Tar-Atanamir that the King should yield the sceptre to his successor before he died'; there were thus a number of years (recorded in the entries of The Line of Elros) between the king's surrender of the sceptre and his death.

5. With this sentence cf. the original version of the Akallabeth, p. 143, $3.

6. It is notable that here and subsequently Avallon is still the name of the whole Isle of Eressea, as it was in the original manuscript A of the Akallabeth, although the later form Avallone and the later meaning (the Haven) entered before that manuscript was completed (see p. 146, $12).

7. Cf. the preamble to the Tale of Years in Appendix B: 'In the Fourth Age the earlier ages were often called the Elder Days; but that name was properly given only to the days before the casting out of Morgoth.' In the Akallabeth 'the Elder Days' was apparently used of the earlier part of the Second Age (p. 156, $53).

8. For other references to the abandoned idea that Gil-galad was the son of Felagund see XI.242 - 3, and pp. 349 - 50.

9. It looks as if the added passage concerning the Dwarves was rejected and replaced immediately. It is strange that my father should have written first that Durin founded Moria at the beginning of the Second Age, with 'his folk' coming from the ruins of Nogrod and Belegost.

10. With this entry compare the headnote to the Second Age in Appendix B. - The words 'the Lady Galadriel of the Noldor, sister of Gil-galad' were not, as might be thought, a slip, but record a stage in her entry into the legends of the First Age. In one of the earliest texts of the work Of the Rings of Power and the Third Age my father wrote of Galadriel: 'A Queen she was and lady of the woodland elves, yet she was herself of the Noldor and had come from Beleriand in the days of the Exile.' To this he added subsequently: 'For it is said by some that she was a hand-maid of Melian the Immortal in the realm of Doriath'; but striking this out at once he substituted: 'For it is said by some that she was a daughter of Felagund the Fair and escaped from Nargothrond in the day of its destruction.' In the following text this was changed to read: 'And some have said that she was the daughter of Felagund the Fair and fled from Nargothrond before its fall, and passed over the Mountains into Eriador ere the coming of Fionwe'; this in turn was altered to: 'For she was the daughter of Felagund the Fair and the elder sister of Gil-galad, though seldom had they met, for ere Nargothrond was made or Felagund was driven from Dorthonion, she passed east over the mountains and forsook Beleriand, and first of all the Noldor came to the inner lands; and too late she heard the summons of Fionwe.' - In the Annals of Aman and the Grey Annals she had become, as she remained, the sister of Felagund.

11. In the Akallabeth the Elendili dwelt mostly in the west of Numenor, and were forced to remove into the east (p. 152); but the statement here that they dwelt mainly in the east may be due simply to compression.

12. This is the first reference to the establishment of a Numenorean settlement at Umbar before the landing of Ar-Pharazon (see p. 156, $41).

13. On the name Valandil for Amandil (as in the first version of the Akallabeth) see p. 156, $44.

14. It is curious that all the texts of the Akallabeth have twelve ships, and only on the late amanuensis typescript did my father change the number to nine (see p. 157, $80); whereas in the present text T 4, certainly no later than the earliest text of the Akallabeth, the number is nine as first written.

15. The statement in this entry concerning the division of the palantiri appeared first in additions to the preceding text T 3; and there they are called Gwahaedir, while the Tower Hills are called Emyn Gwahaedir, replaced by Emyn Hen Dunadan, and then again by Emyn Beraid. This last name does not appear in the actual narrative of The Lord of the Rings.

16. This was probably the first appearance of Amon Amarth, which only occurs in Appendix A (I, i, at end, RK p. 317).

17. All the material in these last entries first appears as rough and complex marginal additions to the manuscript T 3, but at this point there is an addition in T 3 which my father did not take up, perhaps because he missed it:

The shards of the Sword of Elendil are brought to Valandil Isildur's heir at Imladris. He becomes king of the North Kingdom of Arnor, and dwells at Fornost.

The name Valandil of Isildur's heir thus does not appear in T 4; but the entry for 3310 was not added to T 3, and thus Valandil as the name of Elendil's father does not appear in that text.

18. On the ending of the Second Age with the death of Isildur and the loss of the One Ring in the Anduin see p. 170.

19. In Appendix B the entry for S.A.2251 begins 'Tar-Atanamir takes the sceptre. Rebellion and division of the Numenoreans begins.'

In Unfinished Tales (p. 226, note 10) I discussed this, concluding that the entry was certainly an error, although at that time I was apparently unaware of the present text, or at any rate did not consult it,. I suggested that the correct reading should be: 2251

Death of Tar-Atanamir. Tar-Ancalimon takes the sceptre. Rebellion and division of the Numenoreans begins.' No further text is extant before the final typescript from which Appendix B was printed, and it cannot be said how the error arose, moving from

'2060-2251 Reign of Tar-Atanamir' to '2251 Tar-Atanamir takes the sceptre'.

20. I have found nothing in the correspondence of that 'time touching on Appendix A, and I cannot answer the question how it was possible, if the Tale of Years had to be so contracted for reasons of space, to include a further long section in that Appendix at that stage.

21. 'having lived one hundred and ninety-two years': from 3118 to 3310. In the text T 4 3118 was the year of his accession, corrected in the later revision of the typescript T 5 (p. 178) to the year of his birth.

22. The date of Amandil's voyage is given in this text, 3316; it was added also in the revision of the typescript T 5, entry 3310.

23. This appears to be the sole reference in any text to Tolfalas, apart from a mention of its capture by Men of the South in an outline made in the course of the writing of The Two Towers (VII.435).

The isle and its name appeared already on the First Map of Middle-earth (VH.298, 308), but on all maps its extent appears much greater than in the description of it here.

24. On the extremely difficult question of the relation between the destruction caused in Middle-earth in the Great Battle at the end of the First Age, and that caused by the Drowning of Numenor, see V.22 - 3, 32 - 3, 153-4.

VII.

THE HEIRS OF ELENDIL.

While the development of the Appendices as a whole, and the Prologue, was to some degree an interconnected work, the Tale of Years was of its nature (since chronology became a paramount concern of my father's) closely interwoven with the evolution of the history of Numenor and the Numenorean kingdoms in Middle-earth, as has been seen already in the relation of the Tale of Years of the Second Age to the development of the Akallabeth. For the history and chronology of the Realms in Exile the primary document is a substantial work entitled The Heirs of Elendil.

The textual history of this is not easy to fathom. It is divided into two parts, the Northern Line (the Kings and the Chieftains) and the Southern Line (the Kings and the Stewards). The oldest manuscript, which I will call A, is headed The Heirs of Elendil The Southern Line of Gondor; it is clearly if rapidly written for the most part, but in the concluding section recounting the names and dates of the Stewards of Gondor becomes very rough and is obviously in the first stage of composition.

The second manuscript, B, has both the Northern and the Southern Lines, in that order; but though my father fastened the two sections together, they are distinct in appearance. I believe that the second part began as a fair copy of A, but quickly developed and expanded into a much fuller (and increasingly rough) text. To this he added the Northern Line. This section in B seems to be in the first stage of composition (a rejected page shows the names of the later kings and chieftains in the process of emergence) - and there is no trace of any earlier work on the Northern Line, a companion text to A. On the other hand there are clear indications that the Northern Line and its history did already exist when A was set down.

Heavily emended, the composite text B paved the way for a fine manuscript, C; this in turn was much emended in the Northern Line, less so in the remainder, and an amanuensis typescript D was made (much later) from the corrected text (see p. 190).

There is as usual no hint or trace of external dating for any of this work on The Heirs of Elendil, and the most that can be done is to try to relate it to other texts. The relative date of B is shown by the fact that the North Kingdom was still called Arthedain and that Anarion was still the elder son of Elendil, for this was also the case in the third text of the Tale of Years, T 3 (p. 172). The name of the tenth king of the Northern Line is in B Earendil, which is found in the early texts F

1, F 2 of the Appendix on Languages as that of the tenth king (p. 32, footnote to $9). In the fourth text T 4 of the Tale of Years the name of the realm is Amor, Isildur is the elder son, and King Earendur enters.

There can be no doubt therefore that all the fundamental structure and chronology of the Realms in Exile reached written form in the first phase of the work on what would become the Appendices (cf. p. 177).

That the final text C, and many at least of the corrections and additions made to it, belongs to the same time is equally clear. One might suppose this to be the case on general grounds: from the care and calm that are evident in the fine manuscript as it was originally made, in contrast to the latter ragged and chaotic work on the Appendices, and from the fact that corrections to the preceding text B were made (according to my father's constant practice) in preparation for this further version. But the occurrence on the first page of C of the names Valandil of Elendil's father and of Avallon for Eressea (the latter remaining uncorrected) shows that it belongs to the time when the original text of the Akallabeth still stood and T 4 of the Tale of Years had not yet been revised, for both of these have Valandil (pp. 156, 175) and Avallon (p. 173 and note 6). To this may be added the use of

'Noldorin' for 'Sindarin'.

Work on The Heirs of Elendil gave rise to alterations in the text of The Lord of the Rings. A good example of this is found in the passage of the chapter A Knife in the Dark (FR p. 197) where Strider speaks of the history of Weathertop. As this passage stood at the end of work on the chapter (scarcely differing from the original text, VI.169) he said:

There is no barrow on Weathertop, nor on any of these hills. The Men of the West did not live here. I do not know who made this path, nor how long ago, but it was made to provide a road that could be defended, from the north to the foot of Weathertop; some say that Gil-galad and Elendil made a fort and a strong place here in the ancient days, when they marched into the East.

This was altered and expanded, in a late typescript, to the passage in FR, where Strider's account of the great tower of Amon Sul that was burned and broken derives from the addition made to the entry for Arveleg I (eighteenth king of the Northern Line) in Heirs of Elendil B, reappearing in the final text C (see pp. 194, 209). But the addition made to C in the entry for Argeleb I, seventeenth king, 'Argeleb fortifies the Weather Hills', belongs with the alteration of Strider's words about the path, which now became:

The Men of the West did not live here; though in their latter days they defended the hills for a while against the evil that came out of Angmar. This path was made to serve the forts along the walls.

The date of the making of the typescript D, however, is very much later. It is a good text, in top copy and carbon, made by an experienced typist, which fact alone would strongly suggest that it comes from the time after the publication of The Lord of the Rings; but in addition it was made on the same machine as that used for the Annals of Aman, the Grey Annals, the text LQ 2 of the Quenta Silmarillion, and the Akallabeth, about 1958 (see pp. 141-2). It is remarkable (seeing that all the essential material of C had been taken up into Appendix A, if presented there in a totally different form) that my father should have selected this text as one of those to be copied 'as a necessary preliminary to "remoulding" [of The Silmarillion]', as he said in his letter to Rayner Unwin of December 1957 (X.141). He did indeed make use of it later still, writing on the folded newspaper that contains the texts of The Heirs of Elendil 'Partly revised August 1965' - i.e. in preparation for the Second Edition of The Lord of the Rings published in 1966: from this time comes a long insertion in typescript greatly expanding the account of the events leading to the Kin-strife in Gondor, which in somewhat contracted form was introduced into Appendix A in the Second Edition (see further p. 259).