Talking about cave temples, one will naturally recall those of India such as Ajanta and Ellora. Their technique was introduced to China (Tun-huang, Yun-kang and others) along with Buddhism. Compared with Indian cave temples, Chinese ones have many more wall and ceiling paintings, though these are not so characteristic from the architectural point of view. In Pakistan, cave temples are few, but in Afghanistan, we can find interesting western style carved ceilings in dome and laternendecke at Bamiyan etc.

The cave temples in India have no superior in the world in their magnificent carvings and architectural formality, with pillars and beams in order. In contrast to that, the wooden buildings of the ancient times were almost lost, still ca. 1,200 ancient cave temples exist mainly in the Decan Plateau, because they are carved buildings in the strong rocky mountains. About 75 percent of them belong to Buddhism, and it has been considered that they inform us about the figure and construction as replicas of freestanding wooden monasteries and chapels in those times.

A Hindu temple is fundamentally a "House of God," but the Buddhist caves are a combination of "Vihara" caves where priests lived and "Chaitya" caves as chapels enshrining "Stupa." Those caves follow the same manner of the monasteries and chaitya halls built by wood or brick on the ground. We can see the plans of wooden temples by excavated foundations, but the upper building structure is not clear. The ancient wooden architecture consequently was to be analogized from the figure of cave temples.
In fact, Indian cave temples seem to follow the wooden structure in details, and the art and architectural historians like Percy Brown took it for granted that the ancient wooden temples were built as just the same figures as cave temples.

___
Left : Facade of a Chaitya cave, Cave 9 at Ajanta ( India )
Right : Conjectural reconstrucion of wooden building
as an original form of Chaityagriha by Percy Brown

When I wrote the books "The Guide to the Architecture of the Indian Subcontinent" and "Architecture in India," there was nothing for it but to summarize such a common view. It is no problem to think Vihara cave is a replica of a wooden building, being single-story with a flat roof, except that the members are quite thick. However, regarding chaitya cave, it was rather doubtful for me to accept that view for a long time. Chaitya cave possesses a facade with a large window with a pointed arch and interior space looking like a "rafter structure" with a barrel shaped ceiling. It is quite unnatural to consider these elements belong to wooden architecture.
As we can see in the wall paintings like Cave 17 at Ajanta, it is most natural that the wooden building was constructed with a "triangular gabled roof." Such a gabled roof is adopted at any place in the world where rain falls, such as Japanese shrines and Greek temples as well.

___
Left : Cave 26 looking like a "rafter structure", Ajanta (India)
Right : Wall painting showing a wooden house, Ajanta (India)

On the other hand, it is not easy to build a barrel vault like semi cylindrical roof with wooden material. How on earth could they make semicircular rafter as seen in cave temples ? It is impossible to bend big pieces of wood with ancient technology.
The ceilings of early chaitya caves like Baja or Karli are attached by rafters of teak timber so as to look like real wooden buildings, but actually, pieces of board are linked together by "butt joint" at those rafters forming great archs without either metal fittings or bonding agents. If such a rafter structure is applied to a freestanding house without a tie rod, it will easily be destroyed by earthquake or cyclone. In fact, there exists no such wooden building in India.

The art historian, E.B. Havell insisted that the Indian arch was made of bamboo. However, bamboo is also an extremely solid material with a pipe structure with flanges, which is difficult to bend. Apart from slender bamboo used for craft works, they could not bend large pieces of bamboo such as to be used as the building's framework. Even if slender bamboo is used, they should tie up the bent bamboo by a tension member to keep the arch form. However, the arch rafters in chaitya caves are extended freely above the pillars without any expression of a tension member.

After careful consideration, I came to a conclusion as follows; the interior space of barrel vault like ceiling of chaitya caves was created as only an interior design, apart from the fundamental structure of the whole building. The building itself would have been a box type with a flat or gabled roof. Inside of the building, they hung a cylindrical ceiling with wooden arch rafter, made by pieces of board linked by "butt joint" as an interior design. Such a ceiling matches the hemisphere form of stupa, and as we see in the chaitya cave at Karli, they created such a splendid interior space, in harmony with the gentle curve of the stupa.

Interior space of the Chaitya cave, Karli ( India )

The next question is how they got such an extraordinary idea as to compose a cylindrical ceiling by setting arch rafters in a row in the building. It is unlikely to consider that they designed such a form without any model. A more curious point is in the design of the "chaitya window" of the facade. It consists of a large "chaitya arch," and is also carved as if it were a wooden structure; purlins on the rafters are supporting a cylindrical roof. Besides, the whole facade is designed as a sort of pointed arch with a horn on top.
It was in Islamic architecture during 8th -9th centuries that the pointed arch first appeared, and the arches before that period were almost all semicircular shape. If so, why did they have a pointed arch in ancient India ? In addition, why did they carve it as a wooden arch, not the "real arch" of masonry such as stone or brick ? These problems worried me for a long time, and finally I hit upon Lycian cave tombs and sarcophagi in Anatolia (Turkey).

There are caves and sarcophagi with pointed arches in Lycia, moreover carved as if they were wooden structures. Many of them were made in the 4th century B.C. As to India, the first cave temples appeared in the middle of the 3rd century B.C. They are the caves at Barabar and Nagarjuni Hills built for Ajivikas by King Ashoka.
If there is no connection between the two sites located so far apart, it might be considered only a strange coincidence. However, there exists historical evidence of the eastern expedition by Alexander the Great of Macedonia (reign 336 B.C. - 323 B.C.), in the latter half of the 4th century B.C.

Map of Anatolia (Turkey)

According to "Anabasis (The Campaigns of Alexander the Great)" by Flavius Arrianus, Lycians surrendered without fighting and accepted being put under the control of Alexander's army. Nothing was consequently destroyed in Lycia. The book also says a Lycian interpreter-commander accompanied Alexander the Great as far as the Gandhara region in India (present day Pakistan), taking command of the army. It is not difficult to consider that the Lycian architecture was introduced to India together with Greek culture at the time, or after the coming Seleukos Dynasty and Baktria Kingdom.
As the Buddhist images at Gandhara were created by the combination of Greek and Indian sculptures, it is reasonable to imagine Buddhist chaitya caves were created by the combination of Lycian and Indian architectures.

I set up the hypothesis mentioned above based on books of Lycian archaeology and read it at the Summer Seminar of "The Indian Archaeology Society" in summer 2000, and also at "The World Archeology Excavation Academy" in autumn of the same year.
However there was little photo data material, and I had to rely on the etching illustrations in the old books of Charles Fellows, Georges Perrot and so on. There were doubtful points in details, and I had been thinking I should visit Lycia to investigate. In March 2000, I had a chance to visit Lycia, at last, where I could take plenty of photos in many places along with Phrygia and Caria.

Lydian Artemis Temple, Sardis

Recently, Turkey has become a very comfortable country for tourists compared with my first visit of twenty years ago, convenient traffic, comfortable accommodation, delicious and cheap meals, etc. Besides, Turkish people are generally kind, and I could enjoy a pleasant journey. As ‚”he modernization is proceeding also in India since it adopted an open economy policy, a similar comfortable journey might be expected in the near future.
This time again, I visited Istanbul, Edirne, and many places to see Islamic and Byzantine architecture, but my first visit was to Lycia, the main purpose of this journey. From Istanbul I took a night bus to Salihli located near the remains of Sardis, the capital of the ancient Lydian Kingdom.

Lydia reigned the southwest region of Anatolia in the 8th - 6th century B.C., fighting with Phrygia of the central region. Nevertheless, it was conquered by Persia in 546 B.C. and lost her capital Sardis, and later became a Roman territory.
The ruins are left in Sart, but they are mainly of the Roman era, such as a bath, a gymnasium, a synagogue, and a large temple dedicated to Artemis 1 km ahead. However, I was not so interested in Roman ruins this time.

I asked the keeper of the Temple of Artemis about the location of the Necropolis of the Lydian era and went up the hill looking for the vestiges, but regrettably I could not find any. As far as the information of the excavation report of Sardis, which I bought at the superintendent's office of the Gymnasium, it seemed there were rather few remains to see as plastic arts in this site.
According to the report, the tombs which remain here are from after the 6th century B.C. which means a later period than Phrygia, and they were in use through the Persian and Roman eras like in Phrygia and Lycia. The forms of the tombs are both small rock cave and masonry construction with ashlars, with benches for laying down corpses or a chamber to house a sarcophagus. Although it is done in the same way as in Phrygia and Lycia, no relief in facade is remaining here.

Leaving Sart, I dropped by a famous Roman site, Ephesus, and then headed to Lycian sites. However, I will pick up Phrygian ruins first according to the historical order. Nevertheless I visited them later.
The capital of the ancient Phrygian Kingdom is Gordion, about 100 km west from Ankara, but the site was destroyed many times, and there are few things to see left in this site.
The architectural remains in Phrygia are scattered in the hill country between Eskisehir and Afyon. They are rock caves and rock carvings from the latter half of the 8th century B.C. till the first half of the 7th century B.C., which means about 500 years before the first cave temples in India appeared, and they were to be related to Lycian caves in the 4th century B.C.

Phlygian Tomb of Midas, Midas Sehri

At first, I visited Midas Sehri where there was a large town on the hill in ancient times. So it is called Midas Town in English. Now the remains are only a cult throne, tombs and a deep cistern, but a huge rock cut tomb is standing at the foot of the hill. This is the famous Midas Tomb. (The name of King Midas who is famous for the episode of "King's ears are donkey's ears" was likely popular at the time and also in Gordion we can see an old mound called the Tomb of Midas.)

The Tomb at Midas Sehri is a rock cut tomb which is 17 meters high, and has a carved temple facade with a gabled roof. A niche excavated on the lower part is probably for the Goddess Cybele, and the frontispiece is wholly chiseled into geometric shapes. It looks like later Islamic architecture or Modern design. As the general motif of ancient carving is animals or plants, it is surprising to find plenty of abstract patterns on rock surfaces around Midas Sehri. It might have some relationship with the oldest style of Greek art, the so-called "Geometric Style."

Going south from Midas Town, I looked around many remains of Phrygia such as in Bakshish and Yapuldak. It was very difficult to find those remains because they are not the subjects for sight-seeing, small in scale, seriously weathered by such a long term of 2800 years, and besides they are scattered and remote from the villages.
I carried the book "History of Art in Phrygia, Lydia, Caria and Lycia" written by George Perrot and Charles Chipiez, published in London (English edition) more than 100 years ago as one of the series on the art of the Near East. Without the book, surely it would have been impossible for me to research the remains of Phrygia in detail.

___
Phrygian Cave Tombs, Bakshish and Yapuldak

I found that they were basically carved in the shape of houses with a triangular gabled roof and its developed form, temple-type, like the Midas Tomb. In the 8th century B.C. temples in Greece were still built of wood. It seems in Phrygia as well that similar wooden temples like these rock-cut tombs might have been built.
The cave tomb at Bakshish is the most simple in shape, with a relief in a geometric pattern like the Midas Tomb. The small cave tomb of Yapuldak is supposed to be influenced by Greek temples. It has a ridge ornament on the top of the steep gabled roof, and the butt ends imitating wooden purlins are carved along the gable. The columns resemble Ionic order, so its date is possibly later.

In Aslantas nearby, the figure sculptures of two lions facing each other are carved on the rock, just like the "Lion Gate" in Mykenai, showing us the link of civilizations. In Phrygia since then, the lion motif appears frequently, mainly on the pediment.

Temple-type rock tombs of Caria, Kaunos

"Id", the two largest festivals of Islam are referred to as "Bayram" in Turkish. The festival of the breaking fast (Ramadan) is called "Kucuk Bayram," and the festival of sacrifice held in the pilgrimage month is called "Kurban Bayram." The period of the latter of this year (2000) was March 15 -18, falling on my stay in Turkey. During the festival, government offices and companies are all closed, and people living in cities go back to their homelands, just like Japanese the "Bon" festival. The main Turkish transport, long-distance buses and hotels were all full, and on 16th, the historic sites were also closed unfortunately.

Thanks to it I missed the Crusader castle In Bodrum, and I could only enter the Mausoleion of Halikarnassos. However the famous mausoleum of Mausolos, a satrap of Caria was no more than a heap of rubble, which did not let me imagine the ancient majestic figure.
The Carian Kingdom was located in the west of Lycia, and they contended for leadership with each other. However, as Lycia was a more civilized country, many Lycian style rock-cut tombs are found in Necropolises in Caria. The best-preserved are the Kaunos remains located on the opposite bank of the river in Dalyan.

The tombs are carved on the precipitous cliff presupposing nobody enters, so it is better be called cliff-cut tombs rather than rock-cave tombs. They must have been carved by people using rope from the top of the cliff.
It is surprising for one who visits Lycia for the first time, to find the shape of the cliff-cut tombs of Kaunos look like the Greek temples, and also it is surprising that they are arranged in a row like rock cave temples at Ajanta in India. The largest tomb is incomplete and only the upper parts of the four columns are carved, and it proves that in Lycia, they started to carve from upper part just like the Indian rock caves.
The capitals supporting the gabled roof are of Ionian style, which is not surprising because Caria was adjacent to the Ionian Kingdom which made the famous town of Ephesus.

Lycian Tomb of Amyntas in Telmessos

There are many temple-type cave tombs in Lycia, and the best-preserved are in Fethiye, called Telmessos in ancient times. The town was destroyed by an earthquake, and has been restored as a new port town. In the craggy mountains behind the town, many rock-cut tombs are remaining. Mainly they are of house-type rock-cut tombs, and some large temple-type tombs are mixed.
On the tomb at the top of the mountain, the inscription of "Tomb of Amyntas, son of Hermapias" is engraved. Behind the porch with two columns of the Ionic order is a carved portal, and the inside space is a funeral chamber with benches around to lay out the corpse.

In the house type tombs, the facade is carved as if wooden logs are set out as the roof, whereas in the temple-type tombs, square joists are carved, not along the rafter of the roof, but on the horizontal beam. Such a style is common with the Greek temples, and different from the pointed arch-type rock-cut tombs or sarcophagi in Lycia. What would be the reason for such a difference, I wonder ?

There was a bad thunderstorm in Fethiye, but the next day was very fine, and I hired a car to visit various Lycian remains. At first I enjoyed various forms at Tlos under the shining morning sun, such as ruins of the castle, cave tombs, sarcophagi, a Roman theater and a bath. Later, I visited the remains of Pinara, which are spread over vast area, Xanthos having been the ancient capital of Lycia, the sunken city of Kekova in the sea, the flourishing port town of Antiphellos, now called Kas, etc. These Lycian cities which have existed from the 4th century B.C. are scattered along the beautiful coastline of the Mediterranean Sea.

Charles Fellows
(from "Xanthus, Travels of Discovery in Turkey" by Enid Slatter)

It was Charles Fellows (1799-1860), the English archaeologist, who investigated the ancient Lycian cultural remains thoroughly for the first time. His first exploration of Lycia was in 1838, and he wrote a book "A Journal Written during an Excursion in Asia Minor." His second exploration in 1840 was written as a book "An account of discoveries in Lycia." These two books attracted considerable attention from English artists, art historians and archaeologists.
He is nine years older than James Fergusson, an architectural historian who systematized the history of Indian architecture for the first time, and fifteen years older than Alexander Cunningham who is called the father of the Indian archeology. To sum up, these three are nearly in the same generation.

Lycian Sarcophagus and Pillar tomb, Xanthos

In 1848, about 150 years ago, the "Lycian Room" was opened in the British Museum, and the display became very popular. However, in the 20th century the Lycian art and the name of Fellows came to be forgotten gradually, and the items in the Lycian room were scattered. Since then, there is rather little development in Lycian archaeology.
There are many places of interest in the Lycian region, so I will discuss the important architectural remains in accordance with its relation with Indian cave temples.

___
Necropolis at Myra

Although called the "Lycian Kingdom," its actual form in the 4th century B.C. seems to have been a union of city-states, named the "Lycian League." According to Plinius, the league at the time consisted of thirty-six cities, the largest of which was Xanthos. As Lycians were very independent, their region was the last in Anatolia to be placed under the command of the Roman Empire.
There remain no ancient houses or palaces in Lycia, we can see only the castle ruins on the Acropolis, amphitheaters in the Roman age, church ruins in the Byzantine age and so on. The characteristics of the Lycian architecture are most well preserved in Necropolises (city of the dead) built along the slopes of the hills.

The tombs are divided into two categories; i.e. cave tombs and sarcophagi, and the former are divided further into three types. The first one is the "temple-type" that is most monumental (already mentioned in the former chapter). This type is not limited only to Lycia. It is the same type as the wooden or stone temples with triangular gabled roofs in Ionia or Greece, though they are not cave temples.
The second type is the "house-type" tomb and the third is the "pointed arch-type" cave tomb (this will be discussed later). The house-type tombs are seen in necropolises of every city. Some are cliff-cut tombs in the steep rocky mountains, and are difficult to reach, and others are rock caves near the ground level. A lot of house-type tombs are gathered en mass and can be seen at Myra.

The ruins of Myra are located 2km from the town of Demre where the Byzantine church of Saint Nicholas (who is regarded as the origin of Santa Claus) remains. The tens of rock-cut tombs are piled up in the hill behind a well-preserved Roman theater, and the view is so impressive. Their figures are the reproduction of wooden houses, and their pillars and beams are carved as if put together by wooden "halving joints".
The roof is flat type. The thin logs are placed in a row on the beams on which soils are piled up and rammed down. We can find such types of houses in Lycia even today. It is noticeable that when the tombs become more monumental, the ends of groundsills and beams become more warped, forming a decorative shape, which seems difficult to make by a single timber. It is doubtful if the actual wooden houses were made like that.

Reproduced wooden house in Limyra

It is interesting to find reproduced wooden houses that might be the archetype of the house-type tombs that are built at Limyra and actually inhabited. They closely follow the original, but the decorative warped groundsills and beams are made with jointed wooden pieces, which proves the difficulty of warping timber as expressed in the tombs. Perhaps only for the rock-cave tombs, they might be carved with exaggeration.
The external walls are considered mainly to be stone masonry, but there might be the cases of boarding and brick masonry as well. In any case, the pillars and beams are exposed like the Japanese "shin-kabe" structure.
The strangest element is the cross girders laid in the middle of the ground floor wall as frequently seen in the cave tombs. We find the cut ends of joists in a row on the middle girders, but it is improbable to lay down the real joists at such a position. So in the reproduced houses, they are treated as mere decoration.

___
House-type cave tomb called "Painted Tomb", Myra

In the rear hill of Myra, there are the house-type tombs with many splendid relief carvings. 160 years ago, when Charles Fellows visited there, vivid coloring remained partly, and he named it the "Painted Tomb." Although the lower part of the central pillar of the facade has been demolished, a framework of wooden like "post and beam" structure is well remains in good order. The pillar is not flush with the beams in combining together, and it represents a real wooden structure.

These house-type tombs are equivalent to the Buddhist vihara caves in India where priests lived. The facades of the vihara caves have stone masonry which is in different proportion from the Lycian, however, the interior is the perfect replica of the flat roof type wooden building, with large beams between pillars, small beams joining them, and smaller joists lying on top of them. It shows the same principle as the Lycian wooden architecture.
Nevertheless, it does not mean that there might have been an influential relationship between the two countries. A flat roof wooden structure could be built similarly anywhere in the world, and only the small details are different. The rock-cut tombs in Persia are basically the same composition. However, there is no rock cave superior to the house-type tombs in Lycia in the sense of representing the details of the wooden structure so faithfully.

___
Sarcophagi at Sura and Istanbul Archaeological Museum

Ancient people made coffins for the dead either of wood or stone. Sometimes they were stone coffins to put wooden ones in. In any place where the Roman Empire dominated, they used the sarcophagi shaped of gabled houses, and carved splendid reliefs for the noblemen as a memorial.
Such a sarcophagus was inherited from ancient Greece or Anatolia, and basically it consists of two stones; one is a box to lay the corpse in, and another is a lid to cover the box. The popular shape of the lid was a roof of a building, usually a shallow triangular gabled roof. Sometimes the whole sarcophagus looked like a miniature temple building.

However, the Lycians solely developed a different form of sarcophagus. Of the several hundreds of stone sarcophagi that remain in Lycia, some are simple and some are gorgeously carved. The shapes of those lids in Lycia are exclusively of the tall pointed-arched roofs, not the shallow gabled roofs.
In Simena, Cyaenai and Appolonia, a lot of simple and lovely sarcophagi are scattered on the hilltop or along the slope of the hill, and there is a row of ones along the seashore of Theimussa. The house-type sarcophagi with pointed-arched roof scattered in an arbitrary way make us feel as if we had strayed into a wonderland.

Sarcophagi on the hill of Simena

On the roof of a sarcophagus, there are six square prominences; one each to the gable sides and two each to the lateral sides. sarcophagi of higher class people have the lion heads carved on each prominence. The fact that they are limited to lion figure seems to show the tradition from Phrygia. An example is found in a grassy plain at Sura near Demre, but its box part has no ornament.

The most decorative sarcophagus was discovered at Sidon in Lebanon, together with the famous "Alexander Sarcophagus" in the Ionic style. At present, it is displayed at the Archaeological Museum in the site of Topkapi Saray, Istanbul. The gable sides have carvings of the winged angels and Centaurs, and the lateral sides have carvings of the lion heads and cavaliers. On the top, a palmette stands like the Japanese "Oni-gawara."

In the case of the sarcophagus at Sura, a ridgepole is spread at the top, which is the basic type of a Lycian sarcophagus. Checking the gable face (tympanum) carefully, we notice the vertical and horizontal lines are carved surrounding the lion head. These lines show the small posts and beams. In addition, the rafter lines are also carved along the curve of the roof. In short, the Lycian sarcophagi were carved as if they were wooden buildings, like the cave tombs.

___
Upper part of a Lycian sarcophagus in Kas and Tomb of Payava

A sarcophagus preserved in the town of Kas shows the fact more clearly. It stands high in the middle of a road on the high pedestal where inscriptions are engraved in Lycian letters.
The top girders and beams are laid on the pillars of the sarcophagus, small posts supporting small beams stand on the girder, the ends of purlins are carved along the curved rafter, then a ridgepole on the top - it is exactly the same principle as the Buddhist chaitya cave in India, excluding lion heads, that is to say, "the rock-cut pointed arch-type facade that is carved as if it were wooden construction."
The pointed arch-type sarcophagi with such wooden like details are in Pinara, Xanthos and other places. The best example is the "Tomb of Payava," however as already mentioned, it was brought to the British Museum in London in the 19th century, and we cannot see it at the original site.

Why only in Lycia did they make such pointed arch-type sarcophagi, despite the fact that triangular gabled roofs were the ordinary design ? It is difficult to assert the reason exactly, however, we can trace the changes of forms.
Let us check the rock-cut tombs of Myra once again. There are not only the house-type cave tombs with flat roofs, but also the gabled ones. Moreover the ridge of the gable is higher than that of the temple-type, and the slopes of gabled roofs are curved. That is to say, the pointed arch in Lycia is not originated from the "true arch" which is made of laid stones or bricks, but is attained by heightening the ridge and the curving slopes of triangular wooden gable.

Facade of Lomas Rishi Cave at Barabar Hill ( India )

The history of Lycia is almost clear as the ancient epigraphs are deciphered owing to the inscriptions on monuments written in three languages of Greece, Lycian and Aramaic together. According to the inscriptions, the majority of cave tombs and sarcophagi in Lycia were made in the 4th century B.C. It is quite probable these forms and techniques were brought to India after the eastern expedition of Alexander the Great in the late 4th century B.C., and influenced the Lomas Rishi Cave at Barabar Hill that is the first cave temple with an ornamented facade in India.

Though there remains no inscription on the Lomas Rishi cave, the plan and interior space suggest that King Ashoka had it made in the middle of the 3rd century B.C. together with Sudama cave nearby. The technique of rock cave itself might have been brought from Persia, not from Lycia.
Only Lomas Rishi Cave has an ornamentally carved facade in the cave group at Barabar, but there is no similar carving inside. As its design is related directly to the facades of Baja and later caves, it is appropriate to consider that the facade of Lomas Rishi was carried out as a "trial" for those of later chaitya caves, perhaps in the 2nd century B.C.

This facade has been regarded as an accurate replica of wooden architecture at that time. However it is never true, rather it is too strange to be a wooden building form. As already stated, it is irrational from the structural point of view. Its fundamental attitude is the manipulation to make the roof shape as a "pointed arch" by bending forcibly, which should have been a triangular gabled roof by nature. If a house is built like that, the eaves, which should protect the wall from rain, will instead drain the rain onto the wall.
It means that there is an intention to realize the image of the "pointed arch" even against the natural formation of a building. Let us examine further the origin of the unnatural wooden form like this in the cave tombs of Lycia.

Cave Tomb with a facade of pointed arch, Pinara

At the facade of the Buddhist chaitya caves in India, the extrados of the arch is cusped with a horn on top like the Lycian sarcophagi, but its intrados form is semicircular, except a few examples like the Barabar cave. The semicircular shape reminds us of a "true arch" made of laid stones or bricks. However, in fact, the inside of the arch is completely carved like a wooden construction with the rafters and purlins.

It must be the result of the fact that they wanted to build a vaulted roof brought from Persia by means of wood, imitating only the form of a "true arch." As the true arches in stone remain in Buddhist stupa at Guldara in Afghanistan, it is evident that the arch structure had entered India already. However, in India, which belonged to the wooden culture zone at that time, it was not necessary to adopt the technique of the arch structure, therefore they imported the semicircular shape only.
However, it is hardly possible to construct such wooden buildings. It is impossible to bend a timber piece into a semicircular arch shape as a main structural member, unless they utilize the modern technology of wooden assembly. In spite of that, the facades of chaitya caves are carved in wooden like cusped arches. In asking why, I have to say that it is because they adopted the method of the Lycian sarcophagi and rock caves.

___
Left : Upper part of the facade imitating wooden sturucture, Pinara
Right : Upper part of the Chaitya cave, Baja ( India )

There are cave tombs formed with pointed arches at many towns in Lycia. The best example to show clearly the construction method of wooden building being a model of such caves is the cave tomb on the hilltop in Pinara.
Although the lower pillar part was demolished, the upper gable part is well preserved. The latter is composed of a large girder, small posts, a small beam, curved rafters, purlins, and a roof board in this order. The roof is supported firmly by the three posts, which are connected by a small beam. We can see this beam divides rafters on either side into two parts, above and below.
Thus a couple of rafters consist of four members and each of them is almost straight, so one can easily get such loose-curved rafters by planing timber. That is to say, the pointed arch-type wooden construction in Lycia was no doubt a rational structure.

On the contrary, in the process of introducing it to India, Indians neglected the girder and small posts and emphasized only the rafters and purlins that compose the outer form. As a result, they carved the chaitya caves as if they were of complete rafter structure.

The Chaitya cave at Baja that belongs to the earliest stage of Indian cave architecture shows definitely the strangeness of semicircular rafters lacking beams and small posts. Conversely speaking, the interior space of a cave temple enshrining a stupa should be designed as an apsidal plan with a vault-like ceiling, without exposing beams and small posts, to match the hemispherical shape of stupa.
In case of a freestanding wooden building, not in the case of rock caves, such a structure is impossible for construction. Therefore it is proper to consider that they built the large square box type chaitya temples, and made vault-like ceilings inside as a mere interior design to enshrine a stupa effectively. The hypothesis that this was mere inerior design is illustrated by the plans of cave 9 at Ajanta of earlier phase, cave 4 at Aurangabad, smaller chaitya caves at Junnar and so on that are made in the rectangular plans, not in the apsidal plans, in spite of enshrining stupas.

_____ Left : Ajanta Cave No.9 ( Rectangular Chaitya Cave) Plan
Right : Facade of Bhuta Lena cave ( No.26), Junnar ( India )

As for the exterior of such a freestanding chaitya temple, the facade of the Bhuta Lena cave in Junnar shows a good example. Whether the facade was made of wood or stone, large pillars and large girders composed a square frame, and a chaitya window was opened in the inner wall of the porch, and the inside was a chapel covered with a wooden barrel shaped ceiling as an interior design.
Looking carefully the facade of chaitya caves at Ajanta, Karly and others, we can notice that they are also carved in such manner as a large square frame surrounded by the large pillars and beams.

The rock caves were produced all through the world. The ancient rock caves in the Middle East are almost all tombs, not to mention Egypt. The rock cave churches began to be made after the 5th century when the Roman Empire was divided into the East and the West as in Cappadocia. On the other hand, Indians did not produce rock caves as tombs but as monasteries and chapels. It is because there was no custom to build a tomb in India, due to the idea of "Samsara (transmigration)" which explains all the living will transmigrate into another living after death.

Here a question comes out. As the Buddhist stupas are the tumuli built on the Buddha's remains divided after his death, they are in a sense, the tombs of Buddha. The stupa became regarded the typical chaitya (worship object or place), so a chaitya cave became a space to enshrine a stupa. In short, a chaitya cave was a kind of a cave tomb. Why in India, where there is no custom to build tombs, were the cave tombs called chaitya caves made in such a number and in such a magnificent form ?
Stupas were worshiped not only in Buddhism, but also in Jainism (although there is no record that stupas were the tombs of Mahavira). The rock caves first produced in India by King Ashoka were made for the Ajivikas. Their hemisphere space of the rear chamber may suggest the relation to stupa.

There may be a possibility that, because of transmission of the cave tombs of the Middle East, Indians might have begun to construct stupas as tombs and chaitya caves to enshrine them. Differing from the Middle East, where interment was dominant, sarcophagi were not necessary in India where cremation was. Consequently, only the rock cave tombs were inherited to enshrine stupas where the Buddha's remains were put, and they might have come to be used as the chapels as well.
If considering so, two rock beds in each priest's rooms in the vihara caves in India seem very similar to the two benches to lie corpse on in the chambers of rock cave tombs in Anatolia. However, at this stage, I cannot yet declare more about the above-mentioned view.

Whether it is a tomb or a chapel, among the rock caves in the world, those with a pointed arch on the facade, moreover carved as if they were wooden structure, exist only in Lycia and India. The chaitya caves in India, influenced by the method of Lycian cave tombs and sarcophagi, developed in a large scale, and created an imposing barrel shaped interior space setting semicircular rafters as if it was totally "rafter structure," so as to match the hemispherical shape of stupa.
The rational wooden structure in ancient Lycia changed to be the fanciful wooden like cave space in India that cannot exist as an actual wooden building. James Fergusson pointed out that Indian art is extremely "wild in human faith or warm in human feeling" opposing "intellectual" art. Surely the same comment could be applied to the case of ancient cave temples.