As regards buildings I must here be very concise. Brandes has already remarked several times that their chief importance consists in the data they afford for Ilindu-Javanese architecture and later Parmentier has called attention to the same thing though so far provisionally. On this account it was considered better to discuss this section of relief-subjects in the architectural part of this monograph, where such data are treated with the architecture of the whole building. I shall merely enumerate the scenes where these buildings chiefly appear. The greater number of them are palaces, temples, reception-halls and such splendid, richly-decorated buildings, with which in many cases the sculptor though he may have been ruled by the general principles of construction and ornament, has often given rein to his fancy and designed monuments of ideal form, such as did not exist in reality. Buildings of this kind are found specially on 024,31, 33, 43, 100, 103, 124, 131, 140, 141, 143, 14.7, 149, 152,155, Ia6, 10, 15, 19, 25, 26,35,37, 53, 54, 55, 61, 62, 64, 73, 81, 109, 112,Ib 1, 11,16,25,29,33,37,50,79,81,82,84,87,91,96,100,106,107, 112, IBa 16,26,346,IBb10,15,16,22,28,33,40,42,45,52,61,68,72,84,93,95,9799,104,110,113,114,119,125,127,IIB7,8,81,II31,38,68,93,105, 112, 114, 126, 127, III 3—7, 11, 13, the temple series that begins with 20, 69, 70, IIIB 67, 88, IVB 36, 38, IV 42 and 60. Besides these there are buildings which are evidently mostly copied from what the sculptor saw around him and are therefore valuable evidence for the architecture of that time, and for their etImological importance would be worth comparing with what is found now-a-days in Java and other parts of the Archipelago. Besides plain pendapa's on O 7, 14 etc and probably some of those in the series above enumerated, in particular gateways and small temples, there are specimens of this kind to be seen on O 30, 47, 65, 119, 123, 158, Ia 23, 116, Ib 54, 86, IF3a 271, IBb 106, IIIB 87, IVB 4, 14, 18, 21, 50. What a courtyard looked like, we see something of on HI I 1; on the left of this scene is a palatial-looking house, in the centre is a pendapa where the owner of the estate receives his guests and on the right is the rice-shed. Between the house and pavilion in the background there is a fruit tree. A fruit garden enclosed in a fence, but with no house near it, we find on O 61).

Whenever a scene takes place indoors, the sculptor shows us one hall or apartment between the outside walls, sometimes with a smaller side-apartment or gallery next to it or on both sides; but never anything more, so that the addition is intended probably only to explain the situation, not as a section of the building as it actually would be.

In this way we do not learn much. Take for example the large hall on If 36, that is divided into three parts by pillars, each part being roofed with a double arch ornamented with a design of foliage turned inwards (conventionalized makara's), hanging flowers and bells; it is impossible to imagine this to be the section of any real building. It might be true that the roof is supported by columns (probably wooden ones), but the loose foliage under the roof is practically impossible. How the sculptor arrived at this is quite evident; by dividing the space into three with the pillars, he obtained three niche-shaped compartments in a row; these he treated like the niches we so often find on the temples with the same sort of arch and foliage-ornament. The effect is very fine, the design into "nichcs" is very artistic, but it does not give us the impression of a hall. Let us not blame the artist for this, we have not the least right to expect from him an architecturally correct section of the apartment and he is perfectly justified, as long as he gives the spectator the idea of the incidents taking place in a hall, in giving his further attention to producing an artistic relief. So the spectator must be satisfied with this as a work of art, while he knows it shews him nothing of what such a hall is like in reality.

This is the case not only with this relief which I selected as a very clear demonstration of what is to be found on innumerable others and can be accepted generally for all sections of this kind; even those which I note as possibly being taken from reality we must judge as adapted more or less to the sculptors artistic requirements. To achieve his aim he may often have introduced some detail from his own time and surroundings, it is not likely he would intentionally avoid doing this, but unfortunately we have no criterion by which to test the real or fantastic elements here put before us; except in such cases as the one just discussed where any one can see at a glance the practical impossibility of such kind of ornament or construction. Everything remains in general very uncertain and can only be guessed at, for instance whether decoration of garlands that is so usual with the p~ndapa's, and is also found in the roofs of halls should have actually existed and whether they might be attached to flower shaped ornaments in the way we see on II 37 and others. Another kind of roof decoration is made with semi-circular cushions that hang down as if the whole surface were upholstered; this too is seen in pendapa's (f.i. IV 66 where bells hang from the centre of the cushions) and in halls of audience (see, with others IIB 20, where rosettes are attached to the cushions). Sometimes we see only one of these ornements right in the middle of the apartment (Ib 63, IIB 50). It is quite reasonable to suppose that this sort of decoration existed in reality, but how can we make sure of it?

Then added to this is the restricted space available on the reliefs for the buildings. The figures acting in the scene must have the first place and are therefore generally represented in apartments quite out of proportion to their size, in which if they stood up their heads would touch the roof. The result is that the sculptor can only find room for his persons and the mostneeded of the larger pieces of furniture; with the small ones that take up less space it is another case, as we shall notice later on. There was of course no chance for any interior decoration of the apartments. In Maya's sleeping-room on Ia 13, we see only the recumbent queen with her women and one dish; the famous scene of liakyamuni's awakening in the night on Ia 63, shews only the Bodhisattva on a bench and the sleeping women; there was no room for a proper couch. In the hall of the palace on Ib 74, the band of garlands round the ceiling is all we see of decoration to the apartment, the same with the cushion-ornament we noted in the roof on IIB 20; a large couch and a lamp is all the furniture in the room except some small objects put away under the couch. It is the same everywhere; nowhere is there any decoration on the walls. All the same we can get some idea of what the walls look like on reliefs such as O 32 and 73. There the chief person is seated on a wide throne behind which a beautifully decorated wall can be seen. The rest of both scenes takes place in the open air, so that this cannot be meant for the wall of a room, but at any rate it is a large space that has to be decorated. In the first scene this is done with small pillars on each side of a flat panel under a cornice of garlands and rosettes; the second shews a wall covering in the design of a cross and rosette with four lobes, and a row of round vases above). Though it is quite uncertain if the walls of a room were decorated or even if the whole ornament is not due to the sculptor's fancy, it seems suitable to make mention thereof.