In the First Age, Melkor had two fortresses, first Utumno and then Angband. Utumno was the older and greater fortress, but it was destroyed by the Valar in the War for Sake of the Elves. Later, after the Darkening of Valinor, the Dark Lord (now known to the Eldar as Morgoth) returned to Middle-Earth and ruled from Angband, which had originally been a secondary outpost of Utumno.

Since most of Tolkien's tales of the Elder Days take place in Valinor and Beleriand, the second fortress, Angband, is the more prominently featured of the two. After Melkor's poisonous corruption flows out of Utumno near the beginning of The Silmarillion, bringing an end to the Spring of Arda, very little is said about the Dark Lord's original seat. In contrast, Angband at the territory around it, including the desert Anfauglith before its gates and the mountains of Thangorodrim above it, are among the most important elements of the setting.

The older, stronger fortress, from which Melkor did more damage to the world, is a must less prominent element of the legendarium, and I have sometimes found myself wondering why. Which of these two fortresses did Tolkien envision or write about first? Did he create Utumno as part of the backstory for the War of the Jewels, waged against Angband? Or did he actually develop that part of Morgoth's story in closer to chronological order?

  • 1
    A variant of one of them likely appears in the Book of Lost Tales, so I'd look there first.
    – Spencer
    Oct 12, 2021 at 23:49
  • @Spencer - Some of Tolkien's linguistic works preceded the Book of Lost Tales, and both of these place names are found in that original pre-BoLT phase of linguistic work.
    – ibid
    Oct 13, 2021 at 7:14

2 Answers 2


It is unclear. Both terms appear in one of Tolkien's earliest works on Qenya, and thus existed in 1916 before any of the Lost Tales were even written, but the specific meaning may have been added later, and because Tolkien would usually erase his original drafts, it's hard to figure out exactly when that was added.

Both names existed by 1916, likely without the final meanings

The first stage of Tolkien's writing, covering a period from around 1915-1916, consisted of a few scattered poems, some notes, some illustrations, and an extensive linguistic lexicon known as "The Qenya Lexicon".

Tolkien presumably left the Lexicon in England when being deployed, and so even though he would later continue to extensively revise it, a list he made based on the Lexicon when returning to England helps show which words must have been part of the original 1916 stage. John Garth explains this in Tolkien and the Great War.

JRRT is unlikely to have risked taking the lexicon on active service, and its state circa March 1916 may be broadly surmised by excluding all entries lacking in 'The Poetic and Mythologic Words of Eldarissa', a list copied from it probably soon after he returned to England (Parma Eldalamberon 12, xvii-xxi). Some details appearing only in that list are assumed to postdate the initial lexicon phase, and omitted here.
Tolkien and the Great War - Notes to chapter 6

John Garth then proceeds to provide an extensive summary of the state of the mythology as it existed at this point. Based on this, Tolkien seems to have came up with both of the names you're asking about, but only as descriptions of Hell; the character of Melkor did not yet exist and so it's unlikely that their later conception was already present.

Besides wonders, there are monsters in these pages too: Tevildo the hateful, prince of cats, and Ungwë-Tuita, the Spider of Night, whose webs in dark Ruamōrë Earendel once narrowly escaped. Fentor, lord of dragons, was slain by Ingilmo or by the hero Turambar, who had a mighty sword called Sangahyando, or 'cleaver of throngs' (and who is compared to Sigurðr of Norse myth). But there are other perilous creatures: Angaino ('tormentor') is the name of a giant, while ork means 'monster, ogre, demon'. Raukë also means 'demon' and fandor 'monster'.
The fairies know of Christian tradition with its saints, martyrs, monks, and nuns; they have words for 'grace' and 'blessed', and mystic names for the Trinity. The spirits of mortal men wander outside Valinor in the region of Habbanan, which in the abstract is perhaps manimuinë, Purgatory. But there are various names for hell (Mandos, Eremandos, and Angamandos) and also Utumna, the lower regions of darkness. The souls of the blessed dwell in iluindo beyond the stars.
Tolkien and the Great War - Chapter 6 - "Too Long in Slumber"

As best as I can tell from the commentary of edition of The Qenya Lexicon published in Parma Eldalamberon #12, the following seem to be the earliest surviving states of these two words. (Though even so, these could have been written later, as much of the original 1916 definitions were erased and written over.)

ANGAMANDU = EREMANDU. or (pl.) Hells of Iron

UTUMNA lower regions of gloom and dark (not same as MANDOR a place of torment).
Parma Eldalamberon #12, "The Qenya Lexicon", pages 31 and 99, reconstructed from the commentary

Both names with their final meanings existed by c.1917

Tolkien would later make major revisions to this lexicon in 1917-1920 while writing The Book of Lost Tales, along with a parallel work, The Gnomish Lexicon.

ANGAMANDU = EREMANDU. or Angamandi (pl.) Hells of Iron Gn. Angband(or)
UTUMNA lower regions of gloom and dark in the North, Melko’s first dwelling. (Gn. UDUM.)
Parma Eldalamberon #12, "The Qenya Lexicon", pages 31 and 99

Angbann(in) "The Hells of Iron". Melko's (Belca's) great fortress, after the battle of countless Lamentation, down to the battle of the Twilight Pool. also in form Ambann(in), Amannin.
uduvna, Udum = Utumna. (Q.) a dwelling of Melko.
Parma Eldalamberon #11, "The Gnomish Lexicon", pages 19 and 74

Likewise in The Book of Lost Tales themselves, both of these places appear, but Tolkien was revising these Tales a lot, often erasing the original drafts when writing revisions, and I don't know if it's possible to determine which of the two ideas was used first.

Angband is referred to a few times in some of the earlier written tales, namely The Fall of Gondolin (where it's not referred to by name, but rather by "Hells of Iron" or "Hells of Melko") and The Tale of Tinúviel (referred to as "Angamandi"). However, while both of these tales were written earlier, the original versions were erased and all we have to go on are later revisions that Tolkien made, after writing many of the other tales.

Utumno doesn't play a role until some of the later written tales (as "Utumna"), namely "The Coming of the Valar", "The Chaining of Melko", and "The Tale of the Sun and Moon" (where it's first explicitly shown to be a distinct dwelling-place from Angband/Angamandi).

If one of them came first I would lean to it being Angband, but the evidence is inconclusive, and they may very well have been conceived at the same time.

  • I would go so far as to say that even if the name Angamandi didn't exist in the first version of Tuor, the role it played was almost certainly already in the story, even if merely as the 'fortress of the big bad'. Whereas the entire story and plot arc including Utumna surely dated from later. Oct 15, 2021 at 3:47
  • @DavidRoberts - The name Angamandi doesn't even exist in any of the known versions of The Fall of Gondolin, and while I agree that the "Hells of Iron" is essentially the same thing, this story was written a year after both names had already been used in the Qenya Lexicon.
    – ibid
    Oct 15, 2021 at 7:38

Both Utumna and Angamandi appear in The coming of the Valar and the building of Valinor, in BoLT 1, in that order. But the "Hells of Iron" (i.e. Angband) already appear in Tuor and the exiles of Gondolin, which was written in 1916–17, and Utumna does not appear. Nor does Utumna appear anywhere in BoLT 2, which includes The Tale of Tinúviel and Turambar and the Foalókë, both very old stories.

That said, an old name for Melko in Early Qenya is Utumnas, and Utumna (="lower regions") does appear in the Gnomish Lexicon, as does Uduvrin (="Hell-lord"), and the GL dates from 1917. We also get Udum/Uduvna(= “Dark Pit” or “Hell”) in Gnomish (which is where Sindarin "Udûn" comes from, which is mentioned in Fellowship of the Ring).

So it's not clear that the place of Utumna/Utumno in the story as the first fortress of Melko/Melkor was established around 1917, but I would guess Angamandi/Angband entered first, as the fortress of Melko that the Nodoli thralls escaped from in Tuor and the exiles of Gondolin, and the Tale of Tinúviel. The backstory of the rest of the world seems like it was written after, the whole matter of the Lamps, Utumno etc.

  • Utumna is never glossed as "lower regions" in the Gnomish Lexicon.
    – ibid
    Oct 13, 2021 at 6:05
  • @ibid Aha, it is glossed so in Poetic and Mythological Words of Eldarissa, just not in Gnomish Lexicon: eldamo.org/content/words/word-3293124365.html But you evidently had more extensive references to hand! Oct 13, 2021 at 11:54
  • Here's how it looks in PE#12, page 99.
    – ibid
    Oct 13, 2021 at 16:08
  • @ibid thanks! As you say, it's not 100% clear how early those word entries date. My inclination is that it's when places entered the stories that the OP wanted. I included the GL references to show that the word may well have existed, even if not included in the narrative. Oct 14, 2021 at 5:26
  • 1
    GL wasn't written until 1917, possibly after Tolkien was already using the places in his stories. As I've tried showing in my answer, they first appeared in QL sometime before March 1916 (and thus before all of the stories), though we don't know exactly what Tolkien thought they meant at the time.
    – ibid
    Oct 14, 2021 at 6:05

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.