Here appear also certain relations that survived to the latest form. Thus Lórien and Mandos were from the beginning ‘brethren’, each with his special association, of ‘dreams’ and ‘death’ and Nienna stood from the beginning in a close relationship with them, here as ‘the spouse of Mandos’, though afterwards as the sister of the Fëanturi. The original conception of Nienna was indeed darker and more fearful, a death-goddess in close association with Mandos, than it afterwards became. Ossë’s uncertain relations with Ulmo are seen to go back to the beginnings; but Ulmo’s haughtiness and aloofness subsequently disappeared, at least as a feature of his divine ‘character’ explicitly described. Vána was already the spouse of Oromë, but Oromë was the son of Aulë and (Yavanna) Palórien; in the later evolution of the myths Vána sank down in relation to Nienna, whereas Oromë rose, becoming finally one of the great Valar, the Aratar.
Particularly interesting is the passage concerning the host of lesser spirits who accompanied Aulë and Palúrien, from which one sees how old is the conception of the Eldar as quite dissimilar in essential nature from ‘brownies, fays, pixies, leprawns’, since the Eldar are ‘of the world’ and bound to it, whereas those others are beings from before the world’s making. In the later work there is no trace of any such explanation of the ‘pixie’ element in the world’s population: the Maiar are little referred to, and certainly not said to include such beings as ‘sing amid the grass at morning and chant among the standing corn at eve’.*
Salmar, companion of Ulmo, who has appeared in The Music of the Ainur (p. 58), is now identified with Noldorin, who was mentioned by Vairë in The Cottage of Lost Play (p. 16); such of his story as can be discerned will appear later. Subsequent writings say nothing of him save that he came with Ulmo and made his horns (The Silmarillion p. 40).
In the later development of this narrative there is no mention of Tulkas (or Mandos!) going off to round up Melkor at the very outset of the history of the Valar in Arda. In The Silmarillion we learn rather of the great war between the Valar and Melkor ‘before Arda was full-shaped’, and how it was the coming of Tulkas from ‘the far heaven’ that routed him, so that he fled from Arda and ‘brooded in the outer darkness’.
(ii) The earliest conception of the Western Lands, and the Oceans
The earliest map
In The Cottage of Lost Play the expression ‘Outer Lands’ was used of the lands to the east of the Great Sea, later Middle-earth; this was then changed to ‘Great Lands’ (p. 21). The ‘Outer Lands’ are now defined as the Twilit Isles, Eruman (or Arvalin), and Valinor (p. 68). A curious usage, which often appears in the Lost Tales, is the equation of ‘the world’ with the Great Lands, or with the whole surface of the earth east of the Outer Lands; so the mountains ‘towered mightily between Valinor and the world’ (p. 70), and King Inwë heard ‘the lament of the world’ (p. 16).
It is convenient to reproduce here a map (p. 81), which actually appears in the text of a later tale (that of The Theft of Melko and the Darkening of Valinor). This map, drawn on a manuscript page with the text written round it, is no more than a quick scribble, in soft pencil, now rubbed and faded, and in many features difficult or impossible to interpret. The redrawing is as accurate as I can make it, the only feature lost being some indecipherable letters (beginning with M) preceding the word Ice. I have added the letters a, b, c, etc. to make the discussion easier to follow.
Utumna (later Utumno) is placed in the extreme North, north of the lamp-pillar Ringil; the position of the southern pillar seems from this map to have been still undecided. The square marked a is obviously Valmar, and I take the two dots marked b to be the Two Trees, which are stated later to have been to the north of the city of the Gods. The dot marked c is fairly clearly the domain of Mandos (cf. p. 76, where it is said that Vefántur Mandos and Fui Nienna begged Aulë to delve them a hall ‘beneath the roots of the most cold and northerly of the Mountains of Valinor’);* the dot to the south of this can hardly represent the hall of Makar and Meássë, since it is said (pp. 77–8) that though it was not very far from Mandos it stood ‘upon the confines of the Outer Lands’.
The area which I have marked h is Eruman / Arvalin (which ultimately came to be named Avathar), earlier Habbanan / Harmalin (Harwalin), which are simple alternatives (see p. 79).
Later, in a map of the world made in the 1930s, the western shore of the Great Sea bends in a gentle and regular curve westward from north to south, while the Mountains of Valinor bend in virtually the reverse of the same curve eastward,)(; where the two curves come together at their midpoints are Túna, and Taniquetil. Two areas of land in the shape of elongated Vs thus extend northward and southward from the midpoint, between the Mountains and the Sea, which draw steadily away from each other; and these are named Eruman (to the northward) and Arvalin (to the southward).
In the little primitive map the line of the mountains is already thus, and it is described in the text as ‘a great ring curving westward’ (the curve is westward if the extremities are considered rather than the central portion) But the curve of the coast is different. Unhappily the little map is here very obscure, for there are several lines (marked j) extending northwards from Kôr (marked d), and it is impossible to make out whether marks on them are directions for erasure or whether they represent parallel mountain-chains. But I think that in fact these lines merely represent variant ideas for the curve of the Mountains of Valinor in the north; and I have little doubt that at this time my father had no conception of a region of ‘waste’ north of Kôr and east of the mountains. This interpretation of the map agrees well with what is said in the tale (p. 68): ‘the Shadowy Seas to north of Eruman bend a vast bay inwards, so that waves beat even upon the feet of the great cliffs, and the Mountains stand beside the sea’, and ‘Taniquetil looks from the bay’s head southward across Eruman and northward across the Bay of Faëry’. On this view the name Eruman (later Araman), at first an alternative to Arvalin, was taken over for the northern waste when the plan of the coastal regions became more symmetrical.
It is said in the tale (p. 68) that ‘in that vast water of the West are many smaller lands and isles, ere the lonely seas are found whose waves whisper about the Magic Isles’. The little circles on the map (marked k) are evidently a schematic representation of these archipelagoes (of the Magic Isles more will be told later). The Shadowy Seas, as will emerge more clearly later, were a region of the Great Sea west of Tol Eressëa. The other letters on the map refer to features that have not yet entered the narrative.
In this tale we meet the important cosmological idea of the Three Airs, Vaitya, Ilwë, and Vilna, and of the Outer Ocean, tideless, cold, and ‘thin’. It has been said in The Music of the Ainur (p. 58) that Ulmo dwells in the Outer Ocean and that he gave to Ossë and Onen ‘control of the waves and lesser seas’ he is there called ‘the ancient one of Vai’ (emended from Ulmonan). It is now seen that Ulmonan is the name of his halls in the Outer Ocean, and also that the ‘lesser seas’ controlled by Ossë and Ónen include the Great Sea (p. 68).
There exists a very early and very remarkable drawing, in which the world is seen in section, and is presented as a huge ‘Viking’ ship, with mast arising from the highest point of the Great Lands, single sail on which are the Sun and Moon, sailropes fastened to Taniquetil and to a great mountain in the extreme East, and curved prow (the black marks on the sail are an ink-blot). This drawing was done fairly rapidly in soft pencil on a small sheet; and it is closely associated with the cosmology of the Lost Tales.
I give here a list of the names and words written on the drawing with, so far as possible, their meanings (but without any etymological detail, for which see the Appendix on Names, where names and words occurring only on this drawing are given separate entries).
I Vene Kemen This is clearly the title of the drawing; it might mean ‘The Shape of the Earth’ or ‘The Vessel of the Earth’ (see the Appendix on Names, entry Glorvent).
Nme ‘West’.
Valin
or; Taniquetil (The vast height of Taniquetil, even granting the formalisation of this drawing, is noteworthy: it is described in the tale as being so high that ‘the throngs about westward havens in the lands of Men could be seen therefrom’ (p. 68). Its fantastic height is conveyed in my father’s painting, dating from 1927–8 (Pictures by J. R. R. Tolkien, no. 31).)
Harmalin Earlier name of Arvalin (see p. 79).
i aldas ‘The Trees’ (standing to the west of Taniquetil).
Toros valinoriva Toros is obscure, but in any case the first letter of the first word, if it is a T, is a very uncharacteristic one. The reference seems to be to the Mountains of Valinor.
Tolli Kimpelear These must be the Twilit Isles, but I have found no other occurrence of Kimpelear or anything similar.
Tol Eressëa ‘The Lonely Isle’.
I Tolli Kuruvar ‘The Magic Isles’.
Haloisi Velike ‘The Great Sea’.
Ô ‘The Sea’. (What is the structure at the sea-bottom shown below the name Ô? It must surely be the dwelling of Ossë beneath the Great Sea that is referred to in the next tale (p. 106.)
I Nori Landar Probably means ‘The Great Lands’.
Koivienéni The precursor of Cuiviénen, the Waters of Awakening.
Palisor The land where the Elves awoke.
Sil ‘Moon’.
Ûr ‘Sun’.
Luvier ‘Clouds’.
Oronto ‘East’.
Vaitya, Ilwë, and Vilna appear in the three layers described in the tale (p. 65), and Vilna reappears in the bottom right-hand corner of the drawing. There is nothing said in the Lost Tales to explain this last feature, nor is it at all evident what is represented by the curled lines in the same place (see p. 86).
Ulmonan The halls of Ulmo.
Uin The Great Whale, who appears later in the Tales.
Vai The Outer Ocean.
Neni Erùmear ‘Outermost Waters’= Vai.
It is seen from the drawing that the world floats in and upon Vai. This is indeed how Ulmo himself describes it to the Valar in a later tale (p. 214):
Lo, there is but one Ocean, and that is Vai, for those that Ossë esteemeth as oceans are but seas, waters that lie in the hollows of the rock…In this vast water floateth the wide Earth upheld by the word of Ilúvatar…
In the same passage Ulmo speaks of the islands in the seas, and says that (‘save some few that swim still unfettered’) they ‘stand now like pinnacles from their weedy depths’, as is also well seen in the drawing.
It might seem a plausible idea that there was some connection (physical as well as etymological) between Vai and Vaitya, the outermost of the Three Airs, ‘wrapped dark and sluggish about the world and without it’ (at a later point in the Tales, p. 181, there is a reference to ‘the dark and tenuous realm of Vaitya that is outside all’). In the next ‘phase’ of the mythical cosmology (dating from the 1930s, and very clearly and fully documented and illustrated in a work called Ambarkanta, The Shape of the World) the whole world is contained within Vaiya, a word meaning ‘fold, envelope’ Vaiya ‘is more like to sea below the Earth and more like to air above the Earth’ (which chimes with the description of the waters of Vai (p. 68) as very ‘thin’, so that no boat can sail on them nor fish swim in them, save the enchanted fish of Ulmo and his car); and in Vaiya below the Earth dwells Ulmo. Thus Vaiya is partly a development of Vaitya and partly of Vai.
Now since in the earliest word-list of the Qenya tongue (see the Appendix on Names) both Vaitya (‘the outermost air beyond the world’) and Vai (‘the outer ocean’) are derived from a root vaya- ‘enfold’, and since Vaitya in the present tale is said to be ‘wrapped about the world and without it’, one might think that Vaitya-Vai already in the early cosmology was a continuous enfolding substance, and that the later cosmology, in this point, only makes explicit what was present but unexpressed in the Lost Tales. But there is certainly no actual suggestion of this idea in any early writing; and when we look again at the drawing it seems untenable. For Vai is obviously not continuous with Vaitya; and if the appearance of Vilna in the bottom of the drawing is taken to mean that the Earth, and the ocean Vai in and on which it floats, were contained within the Three Airs, of which we see the reappearance of the innermost (Vilna) below the earth and Vai, then the suggestion that Vaitya—Vai were continuous is still more emphatically confounded.
There remains the baffling question of the representation of the world as a ship. In only one place is there a suggestion that my father conceived the world in such a way: the passage that I have cited above, in which Ulmo addresses the Valar on the subject of Vai, concludes:
O Valar, ye know not all wonders, and many secret things are there beneath the Earth’s dark keel, even where I have my mighty halls of Ulmonan, that ye have never dreamed on.
But in the drawing Ulmonan is not beneath the ship’s keel, it is within the ship’s hull; and I am inclined to think that Ulmo’s words ‘beneath the Earth’s dark keel’ refer to the shape of the Earth itself, which is certainly ship-like. Moreover, close examination of the original drawing strongly suggests to me that the mast and sail, and still more clearly the curved prow, were added afterwards. Can it be that the shape of the Earth and of Vai as he had drawn them—with the appearance of a ship’s hull—prompted my father to add mast, sail, and prow as a jeu d’esprit, without deeper significance? That seems uncharacteristic and unlikely, but I have no other explanation to offer.*
(iii) The Lamps (pp. 69–70)
In this part of the narrative the tale differs remarkably from the later versions. Here there is no mention of the dwelling of the Valar on the Isle of Almaren after the making of the Lamps (The Silmarillion p. 35), nor of course of the return of Melko from ‘outside’—because here Melko not only did not leave the world after entering it, but actually himself made the pillars of the Lamps. In this story, though Melko was distrusted by some, his guileful co-operation (even to the extent of contributing names for the pillars) was accepted, whereas in the later story his hostility and malice were known and manifest to the Valar, even though they did not know of his return to Arda and the building of Utumno until too late. In the present tale there is a tricksiness, a low cunning, in Melko’s behaviour that could not survive (yet the story of his deceitful making of the pillars out of ice survived into the versions of the 1930s).
Later, it was the Lamps themselves that were named (ultimately, after intervening forms had been devised and discarded, Iluin the northern Lamp and Ormal the southern). In The Silmarillion Ringil (containing ring ‘cold’) survived only as the name of Fingolfin’s sword, but Helcar is that of the Inland Sea which ‘stood where aforetime the roots of the mountain of Illuin had been’ (p. 49). In the present tale Helkar was the name of the southern, not the northern, pillar. Now helkar meant ‘utter cold’ (see the Appendix on Names), which shows that Helkar was originally in the extreme south (as it is in one of the two positions given for it on the little map, p. 81), just as Ringil was in the extreme north. In the tale there is no mention of the formation of Inland Seas at the fall of the Lamps; this idea appeared later, but it seems virtually certain that it arose from the story of the melting pillars of ice.
There is no later reference to the building of the Mountains of Valinor from great rocks gathered in Eruman / Arvalin, so that the region became flat and stoneless.
(iv) The Two Trees (pp. 71–3)
This earliest account of the uprising of the Two Trees illuminates some elements of later versions more concentrated in expression. The enduring feature that the ground beneath Silpion (Telperion) was ‘dappled with the shadows of his fluttering leaves’ (The Silmarillion p. 38) is seen to have had its origin in the ‘throbbing of the tree’s heart’. The conception of light as a liquid substance that ‘splashed upon the ground’, that ran in rivers and was poured in cauldrons, though not lost in the published work (pp. 38–9), is here more strongly and physically expressed. Some features were never changed, as the clustered flowers of Laureli
n and the shining edges of its leaves.
On the other hand there are notable differences between this and the later accounts: above all perhaps that Laurelin was in origin the Eldar Tree. The Two Trees had here periods of twelve hours, not as later seven;* and the preparations of the Valar for the birth of the Trees, with all their detail of physical ‘magic’, were afterwards abandoned. The two great ‘cauldrons’ Kulullin and Silindrin survived in the ‘great vats like shining lakes’ in which Varda hoarded ‘the dews of Telperion and the rain that fell from Laurelin’ (ibid. p. 39), though the names disappeared, as did the need to ‘water’ the Trees with the light gathered in the vats or cauldrons—or at any rate it is not mentioned later. Urwen (‘Sun-maiden’) was the forebear of Arien, Maia of the Sun; and Tilion, steersman of the Moon in The Silmarillion, who ‘lay in dreams by the pools of Estë [Lórien’s wife], in Telperion’s flickering beams’, perhaps owes something to the figure of Silmo, whom Lórien loved.