In each niche there lies a small lotus cushion but without any image. Even in 1834 during the digging up of the ruin buried under an overgrown mound, no images were found in- or outside these niches.
What then was the meaning of them?
They were explained to us by the French Indian architect Henry Parmentier who spoke of analogical cases in Farther India [Bulletins de l’école française d’Extrême Orient][13]. Even there the temples closely related to the Hindu ruins in Java had no windows or openings outside the entrance which opened into an equally dark porch; and as it was very dark inside the walls were provided with niches for lamps to light the images throning in these sanctuaries.
After mature consideration I came to the conclusion that the niches of tyanḍi Mĕndut must also have had this destination, and this may be the reason why all of them were affixed in front and opposite (not behind) the three images, so that I never doubted the four walls to have had any other opening than the door which opened through the front wall into the almost equally dark porch.
This conviction of mine has been confirmed by some corresponding cases, among others, by the fact that the four still undamaged walls of the comparatively large inner-rooms of tyanḍi Sévu in the plain of Parambanan, have no other opening but the door which gives entrance to the (eastern) porch. However, we don’t see any niche in the inner-room of tyanḍi Kalasan, perhaps because there was room enough in these two sanctuaries to place one or more lights before or on the altars which carried the Buddha or Târâ image.
In the main temples of the Parambanan group, with the exception of tyanḍi Shiva, there was no place for these lights. The altar-shaped pedestals of the images were much smaller there, and round about them there was but little room.
This temple’s walls hewn with exquisitely modelled festoons had also no niches, and could not have had them unless one would have partly sacrificed its panels. But in all other, less spacious temples whose walls were unadorned, are still to be found simple and square formed stones, 2 of which we see in each side-wall, and 1 on every side of the entrance through the front wall, consequently just as the 6 niches in tyanḍi Mĕndut and equally fit to the same purpose. Had not the front walls of these sanctuaries partly fallen down I am sure we then could see that they also had no windows above the entrances, and that neither the inner-rooms of tyanḍis Sévu and Kalasan, nor the sanctuary of tyanḍi Mĕndut ever had them till before some years when the president of the “Oudheidkundige Commissie” (board of antiquarian science) ordered these openings to be pierced through the front wall scarcely rebuilt by Van de Kamer. And that, contrary to this architect’s official objections, and against my not-official but well argued warning. An irresponsible deforming, a violation of the original architecture, a desecration of a primevally pure style!
And this becomes much clearer to us when we raise our eyes, and fully see how this polygonal hole spoils the harmony of the character of the pyramidical vault so beautifully thought, and which I mean to have once known as a closed whole.
Those who contemplate this pseudo-vault unprejudicedly will no more regret than I do, that such a thing could have happened without having been redressed up to this date. It is true, it would cost much labour again, and money too, but this labour and money would undoubtedly be far better accounted for than that which was uselessly spent to commit such an unpardonable mistake.
Dr. Brandes may have been deceived by the form of the hole the dropping stones had made outside in the front wall above the entrance, and which he knew from engravings only, for, when he first visited this temple Van de Kamer had this wall erected again just as it once was, and without any other opening but the door. On account of analogical Indian ruins pictured in Fournerau’s and Porcher’s works, I stated elsewhere how the falling asunder of such walls which had been run up with hewn stones without mortar, are to form the very same angular lines of breach Dr. Brandes unrightly ascribed to the architect’s intention to build them so.