PREFACE.                               
                                                                       
 This  fifth  volume  of  The  History  of  Middle-earth   completes  the
 presentation  and analysis  of my  father's writings  on the  subject of
 the First Age up to the time at  the end  of 1937  and the  beginning of
 1938  when  he  set  them  for  long  aside. The  book provides  all the
 evidence  known  to  me  for  the  understanding  of his  conceptions in
 many  essential  matters  at the  time when  The Lord  of the  Rings was
 begun;  and  from  the  Annals  of  Valinor,  the  Annals  of Beleriand,
 the  Ainulindale,  and  the  Quenta  Silmarillion  given  here   it  can
 be   quite   closely   determined  which   elements  in   the  published
 Silmarillion  go  back  to  that  time,  and  which  entered afterwards.
 To  make  this a  satisfactory work  of reference  for these  purposes I
 have thought it essential to give the texts of the later 193Os  in their
 entirety,  even  though  in  parts  of the  Annals the  development from
 the  antecedent  versions  was  no'.  great;  for the  curious relations
 between  the  Annals  and   the  Quenta   Silmarillion  are   a  primary
 feature  of  the  history  and here  already appear,  and it  is clearly
 better to have all the  related texts  within the  same covers.  Only in
 the case of  the prose  form of  the tale  of Beren  and Luthien  have I
 not  done  so,  since  that  was  preserved  so  little  changed  in the
 published  Silmarillion;  here  I  have  restricted  myself to  notes on
 the changes that were made editorially.                                
   I cannot,  or at  any rate  I cannot  yet, attempt  the editing  of my
 father's  strictly  or narrowly  linguistic writings,  in view  of their
 extraordinary  complexity  and difficulty;  but I  include in  this book
 the  general   essay  called   The  Lhammas   or  Account   of  Tongues,
 and  also  the  Etymologies,   both  belonging   to  this   period.  The
 latter,  a   kind  of   etymological  dictionary,   provides  historical
 explanations  of  a  very  large   number  of   words  and   names,  and
 enormously   increases   the   known   vocabularies   of    the   Elvish
 tongues  -  as  they were  at that  time, for  like everything  else the
 languages  continued  to  evolve  as  the  years  passed.  Also hitherto
 unknown  except  by  allusion  is  my  father's  abandoned 'time-travel'
 story  The  Lost  Road,  which  leads  primarily  to  Numenor,  but also
 into  the  history  and  legend  of  northern  and western  Europe, with
 the  associated  poems  The  Song of AElfwine  (in  the stanza  of Pearl)
 and  King  Sheave  (in  alliterative  verse).  Closely   connected  with
 The  Lost  Road  were  the  earliest   forms  of   the  legend   of  the

                             
                                                                    
 Drowning  of  Numenor,  which  are  also  included  in the  book, and
 the first glimpses of the  story of  the Last  Alliance of  Elves and
 Men.                                                                
  In  the  inevitable  Appendix I  have placed  three works  which are
 not  given complete:  the Genealogies,  the List  of Names,  and the
 second  'Silmarillion'  Map, all  of which  belong in  their original
 forms  to  the  earlier  193Os.  The Genealogies  only came  to light
 recently,  but they  add in  fact little  to what  is known  from the
 narrative  texts.  The  List   of  Names   might  have   been  better
 included in Vol. IV, but  this was  again a  work of  reference which
 provides  very  little  new  matter,  and it  was more  convenient to
 postpone it and then to give just those few  entries which  offer new
 detail. The  second Map  is a  different case.  This was  my father's
 sole  'Silmarillion'  map  for  some  forty  years,  and here  I have
 redrawn it to show it as it was when first made, leaving out  all the
 layer  upon  layer  of later  accretion and  alteration. The  Tale of
 Years and  the Tale  of Battles,  listed in  title-pages to  The Sil-
 marillion as elements in that work  (see p.  202), are  not included,
 since  they  were  contemporary  with  the   later  Annals   and  add
 nothing  to  the  material  found in  them; subsequent  alteration of
 names and dates was also carried out in a precisely similar way.
  In  places  the detailed  discussion of  dating may  seem excessive,
 but since  the chronology  of my  father's writings,  both 'internal'
 and 'external', is extremely difficult to determine and  the evidence
 full of traps,  and since  the history  can be  very easily  and very
 seriously  falsified  by mistaken  deductions on  this score,  I have
 wished to make as plain as I can the reasons for my assertions.     
  In  some  of  the  texts  I  have   introduced  paragraph-numbering.
 This is done in the belief that it  will provide  a more  precise and
 therefore  quicker  method  of  reference  in a  book where  the dis-
 cussion of its nature moves constantly back and forth.              
  As  in  previous  volumes  I  have   to  some   degree  standardized
 usage in respect of  certain names:  thus for  example I  print Gods,
 Elves,  Orcs,  Middle-earth,  etc.  with  initial capitals,  and Kor,
 Tun,   Earendel,   Numenorean,   etc.   for   frequent    Kor,   Tun,
 Earendel, Numenorean of the manuscripts.                            
  The earlier volumes of the series are referred to as I (The  Book of
 Lost Tales Part I), II (The Book of  Lost Tales  Part 11),  III (The
 Lays   of  Beleriand),   and  IV   (The  Shaping   of  Middle-earth).
 The sixth volume  now in  preparation will  concern the  evolution of
 The Lard of the Rings.                                              
  The  tables  illustrating  The  Lhammas  are  reproduced   with  the

 permission  of the  Bodleian Library,  Oxford, who  kindly supplied
 photographs.                                                       
                                                                   
   I list here for convenience the abbreviations used in the book in
 reference to various works {for a fuller account see pp. 107 - 8).
                                                                   
 Texts in Vol. IV:                                                  
 S. The Sketch of the Mythology or 'earliest Silmarillion'.          
 Q. The Quenta ('Quenta Noldorinwa'), the second version            
 of 'The Silmarillion'.                                             
 AV1. The earliest Annals of Valinor.                               
 AB1. The earliest Annals of Beleriand (in two versions, the        
 second early abandoned).                                           
 Texts in Vol. V:                                                   
 FN. The Fall of Numenor (FN I and FN II referring to the            
 first and second texts).                                           
 AV2. The second version of the Annals of Valinor.                  
 AB2. The second version (or strictly the third) of the Annals      
 of Beleriand.                                                      
 QS. The Quenta Silmarillion, the third version of 'The              
 Silmarillion', nearing completion at the end of 1937.              
 Other works (Ambarkanta, Ainulindale, Lhammas, The Lost             
 Road) are not referred to by abbreviations.                        
                                                                   
   In conclusion, I take this opportunity to notice and  explain the
 erroneous  representation of  the Westward  Extension of  the first
 'Silmarillion'  Map  in  the  previous   volume  (The   Shaping  of
 Middle-earth p.  228). It  will be  seen that  this map  presents a
 strikingly different appearance  from that  of the  Eastward Exten-
 sion  on  p.  231. These  two maps,  being extremely  faint, proved
 impossible   to   reproduce  from   photographs  supplied   by  the
 Bodleian  Library,  and  an  experimental  'reinforcement'  (rather
 than  re-drawing) of  a copy  of the  Westward Extension  was tried
 out. This I rejected, and it was then found that my  photocopies of
 the originals  gave a  result sufficiently  clear for  the purpose.
 Unhappily,  the  rejected  'reinforced'  version  of  the  Westward
 Extension  map  was  substituted  for  the  photocopy. (Photocopies
 were also used  for diagram  III on  p. 247  and map  V on  p. 251,
 where the originals are in faint pencil.)