| articles

Download the PDF version of this article

The Dancing Men of Kanchipuram, an Anomaly in Pallava Land
by Drs. Liesbeth Pankaja Bennink

text and images © the author except where otherwise noted

Published July 30, 2019

(click on the small image for full screen image with captions.)

Fig. 1
Do you recognize this? You walk into a place and something catches your attention, so you quickly take a few shots. But it is not what you came for; you are on your way to a different goal, your actual subject at the time, so you move on quickly, not wanting to waste precious time. Then, having finished what you came for -- sometimes years later -- you realize your quick clicks by the way are revealing something quite remarkable. The photos reveal something raising a ton of questions in your mind, and the need for more photos. But now the subject is halfway around on the other side of our little world, quite out of reach.

I had one such experience during my last visit to India. But I have just enough photos with me to be able to share with you this remarkable find in the field of South Indian sculpture, architecture and history.

Two Forgotten Pillars

The courtyard of the Kachabeshwarar temple in Kanchipuram is a wide-open space with shrines arranged around its periphery, except for the Ganesha shrine which is standing free in the open space by the side of the temple tank. This North-facing shrine is fronted by an open pillared pavilion. At the northern corners of this mandapa we find two unique artifacts standing lost in space, in retrospect seemingly forgotten in plain sight (fig. 1).[1] These two remainders of pillars (I can’t think of a better way to describe them) are, on the one hand clearly Indian in their artistic vocabulary, but on the other hand they don’t fit in with any of the categories usually applied to classify (South) Indian art, such as Pallava, Chola, or Chalukya, after the names of the dynasties that ruled this part of the world at one time or another.

Fig. 2

Fig. 3

Fig. 4

Fig. 5

Fig. 6

Fig. 7
The sculptural program of these two pillar fragments is like nothing I have seen anywhere in the Indian context. Framed between the familiar elements of base and mālāsthāna both pillars represent four men who move or dance around its column. The figures are moving clockwise on the one (figs. 2-4), and anticlockwise on the other pillar (figs. 5-7). The representation depicts them as individuals, not as standardized representations of a class, as you find with the dance freezes from the Chola or Vijayanagara period. Some have beards, some don’t. One of them wears a sacred thread. Their faces are that of individuals with recognizable traits, possibly representing people from various ethnic backgrounds. The men represented appear to be approximately the same age, possibly in their forties, with slightly bulky frames. They all wear the same simple traditional South Indian dhoti with aṅgavastram (a shawl on the upper body) in the same way, and a hairdo or headdress that looks like a Phrygian cap.

Although I did not measure them when I took the photos, the height of the pillar fragments is approximately 120 cm, approximately reaching up to my chest, and maybe some 70 to 80 centimeters in diameter, as estimated from my memory and the photos.[2] The material they are made of looks like the type of granite found in the ranges of the Eastern Ghats around Kanchipuram. The pillars are round all through what remains of their height.

Their Place in Space and Time?

Kanchipuram is primarily associated with the Pallava dynasty for which it served as capital from the 6th to 9th centuries. Its rulers are associated with the great masterpieces of Mamallapuram, the first known stone structural temples in Southern India like the Kailāsanatha temple in Kanchipuram and others, and many excavated cave temples. The Pallava dynasty ruled over an extensive area of South East India from somewhere in the early Christian era till the end of the 9th century, which is when they lost their dominance to the Chola dynasty. Their earlier period is lost in prehistory. During their rule several wars were fought with Chalukya rulers from Badami, with both conquering the other´s capitals.[3]

We can use elements of a pillar or pilaster like the base and the padmabandha with mālāsthāna to attempt to place these pieces in the chronology of the area. Pillars are commonly analyzed on the basis of their cross-section as round, square, eight-sided, and multi-or sixteen-sided and further according to the characteristics of their constituting elements such as base, shaft, mālāsthāna, padmabandha, kalaśa, kumbha, kamala, palagai and podigai.[4]

Padmabandha with Mālāsthāna

Fig. 8
Of the base of these two pillars only a part is visible. It may be that the base has been cut and shortened and the pillars have also been sunk into the pavement. One pillar has been better preserved than the other. Of the less preserved pillar (which I have named as pillar II) only a bit of one molding is still visible. The base of the better-preserved pillar shows (from bottom up) an inverted lotus molding, a round, ring-shaped molding decorated with floral patterns, and an upturned lotus molding (fig. 8). I have not been able to connect this order to any known pillar form either through the literature or in my own photo archive, except that they could be interpreted as somewhat similar to the base of the pillars of the cloister of the Vaikuntha Perumal temple; but in this temple we don’t see the detail of decoration and workmanship.[5]

The padmabandha with mālāsthāna is a common decorative element found on pillars and pilasters[6] in South Indian (Dravida) temple architecture all through the medieval period. The padmabandha is a band of lotus petals circling the pillar shaft. Under this we find the mālāsthāna defined as “the pillar-part below the ghata, decorated with pearl and floral festoons”.[7] The padmabandha and mālāsthāna are always joined together, inseparable, on both round and square pillars and pilasters. There is considerable variation in the artistic details of this element. Within the strict structural framework, it offered a considerable amount of artistic freedom to the sculptors. It can consist of several parallel layers or bands the top one of which is a row of upturned lotus petals, the padma. The lowest layer is a series of entwined garlands or festoons, the mālā.

Fig. 9

Fig. 10

Fig. 11

Fig. 12
Pallava Pillars and their Mālāsthāna

Of pillar II only a part of the mālā survives (figs 11-12). In pillar I we can still see the full mālā and also a fragment of the covering element, which seems to be a floral band (figs 9 -10). The representation of the festoons is executed in deep and bold relief.

As Kanchipuram is the heartland of Pallava art it is logical to first compare these two pillar fragments with pillars from the monuments of this dynasty in order to try and situate them within the chronology of South Indian art history. Pallava excavated cave temples, structural temples and rock carved architecture and art are the earliest known in South India; from earlier periods there are only megalithic remains.

Fig. 13
The cave temple of Mandagapattu is understood to be the earliest remaining architectural creation attributed to a Pallava ruler. Mahendravarman I’s (600-630) earliest excavations have solid and compact pillars divided in a square-octagonal-square lay-out, such as those seen at Mandagapattu (fig.13). It is understood by art historians that Pallava temple architecture evolved from these solid and simple pillar forms to more elegant and complex ones.

The pillars that are especially associated with Pallava temple architecture are round or octagonal with a seated lion or yali (horned lion)[8] as caryatid. Lion and yali caryatids are found in Pallava temples from the time of the Pallava king Narasimhavarman I, also known as Mahamallan (630-668), the successor of Mahendravarman, as in the Mamallapuram Mahishasuramardini cave and Ramanuja mandapam. Pilasters of the same architecture, however, are square. Lion carried pillars are associated with the cave temples and rathas found in Mamallapuram, and with the structural temples built by king Narasimhavarman II, Rajasimha (700-728), the Kailāsanatha temple at Kanchipuram and the Shore Temple at Mamallapuram;  and also temples attributed to later Pallava kings like the Vaikuntha Perumal, the Mukteshvara and Matangeshvara temples, all in Kanchipuram. All the pillars and pilasters of these structures have associated padmabandha with mālāsthāna ornamentation.

The first example is a slim pillar from the porch of the Mahishasuramaridini cave temple in Mamallapuram, attributed to king Narasimhavarman I Mahamalla in the 7th century [9]. As is typical for this period it is sixteen sided with an elegant padmabandha and mālāsthāna. The carving is shallow but shows fine workmanship. The adjacent pilaster has the same decoration in square form.

Fig. 14

Fig. 15

Fig. 16

Fig. 17
The next illustrations are the pillars from the Ganesha Ratha in Mamallapuram (figs 14-17). Also sixteen sided they are not as slim and elegant, but have impressive yalis, horned lions, for caryatids. The padmabandha and mālāsthāna are shallow carved. The pilasters of the same temple have gryphons as caryatids, with the same padmabandha and mālāsthāna decoration.

Also in the Kailāsanatha temple in Kanchipuram the padmabandha and mālāsthāna elements are an important feature of the pillars and pilasters. The example given here is the niche with the relief depicting Shiva as Dakshinamurti. The padmabandha with mālāsthāna element is placed near the top of the square pilasters (figs 18-19). Here the mālāsthāna element is more strongly defined and carved quite a bit deeper than in Mamallapuram. Halfway down there is another padmabandha, but without the mālāsthāna. At the base of the pilasters there is a strongly defined mālāsthāna element emerging from the mouths and claws of miniature yali faces.

Fig. 18

Fig. 19
The next illustrations are from the Parameswara Vinnagaram or Vaikunta Perumal Temple in Kanchipuram, thought to have been built by the Pallava king Nandivarman II (731-796) (figs 20-22). The caryatid lions are seated on a base which is similar in moldings to the bases of the two Dancing Men pillars without showing the clear decoration of lotus petals. An interesting feature is that among the original lions in sandstone there are a few carved from granite. Although the workmanship is at least as excellent as that of the original Pallava caryatids, the details of the decorations of the padmabandha and mālāsthāna are rather different between the sandstone pillars and the granite ones, suggesting these may be from an altogether different workshop and time. Both offer us a glimpse into artistic freedom within the context of highly defined doctrinal concepts of temple architecture. Although the mālāsthāna element is prominently present in both sculptural plans, neither resembles in any way that of the two pillar fragments in the Kachabeshwarar temple. The carving is shallow and the alternating festoons are depicted as chains of pearls.

Fig. 20

Fig. 21

Fig. 22
Within the corpus of Pallava architecture the pillars from the lower cave excavated in the Rock Fort of Tiruchirapalli are our concluding comparative illustrations from the Pallava territory (figs 23, 24). Although this temple cave is clearly an early construction it cannot with certainty be attributed to any of the Pallava kings, because there are no inscriptions. It has been dated to the 8th century. The pillars are octagonal from the square base upwards.The padmabandha with mālāsthāna is well defined. It is carved relatively boldly with well-defined details. The festoons are strings of pearls.

Fig. 23

Fig. 24
From this comparison with mālāsthāna features from various representative examples of Pallava temple architecture we may conclude that the mālāsthāna of the two pillars remnants is the Kachabeshwarar temple stand out as unique in their boldness and well defined carving. From all the given illustrations the mālāsthāna from the Kailāsanatha temple (figs 18 and 19) seem closest in comparison, similar in the boldness and the definition and character of the carving. But the way in which the festoons of the mālāsthāna are shaped is quite different. Their daring energy is not found anywhere else.

The conclusion at this point is that on the basis of the mālāsthāna these pillars could belong to the Pallava era. But at the same time their boldness and brash energy give a very different impression from the controlled and refined mālāsthāna of the many existing monuments from the Pallava era. On the whole I would suggest that this may point to an early date.

Contemporary and later Dynasties

As we look further afield for a possible historical context for these Dancing Men pillars, we can look at Badami (figs. 25-26), Pattadakal (figs. 27-29) and Ellora (figs. 30-31) to compare the mālā element with possible contemporary or connected temple architecture. These monuments are associated with the Chalukya and Rashtrakuta dynasties, the main rival dynasties contemporary with the Pallava reign in Kanchipuram. Several wars and many battles were fought and both the Pallava and Chalukya capitals were conquered at one time or another by the opposite army.

Fig. 25

Fig. 26

Fig. 27
For the most part the pillars from the Chalukya and Rashtrakuta periods are squat and square (figs. 25, 28, 30). The padmabandha and mālāsthāna elements are elegant and elaborate, executed in shallow carving (figs. 26, 29, 31). Contrasting the pillars illustrated from the northern niche of the Galaganatha temple at Pattadakal (fig. 27) have a square base, round shaft and octagonal padmabandha with mālāsthāna. Here too the mālāsthāna and padmabandha are elegant and executed in shallow carving. Still looking a little further into the chronology all throughout the Chola (Early Chola figs. 32, 33; Later Chola figs. 34-35) period up to the Later Pandya (fig. 36) and Vijayanagara period (fig. 37) the padmabandha and mālāsthāna elements are continued as decoration on slim and elegant pillars and pilasters which can be round, square, octagonal or multifaceted (16 sided). Invariably the carving is shallow and the style mostly elaborate and elegant. A few examples are illustrated.

Fig. 28

Fig. 29

Fig. 30

Fig. 31

Fig. 32

Fig. 33

Fig. 34

Fig. 35

Fig. 36

Fig. 37
Dancing Men

Whereas the base and the mālāsthāna place these two pillars solidly within the framework of Dravidian architecture, the sculptural plan of the main reliefs of men arranged around the main body of the column makes them unique. I haven’t seen anything like it anywhere else. Each pillar frame depicts four men moving, possibly dancing, around the body of the column, on the one pillar in clockwise, on the other in anticlockwise direction. On each pillar one of the four is carrying an unidentified round object on the right shoulder.

The eight men are depicted individually and realistically. This is unlike what we see in dance reliefs from the Chola or the Vijayanagara period, where dancers and musicians are depicted as stylized characters reflecting a ´type´. The faces could be portraits. Their body is that of a man in his early to middle forties, with a slight belly. This is very different from the stylized bodies of the Pallava, Chola and Vijayanagara periods, where men are depicted with flat bellies and athletic bodies and are somewhat stylized, with just a slight fold above the waistband formed by the tie of the dhoti. An exception is the depiction of rishis, in which case they have stocky bodies with pot-bellies.

Fig. 38

Fig. 39
The posture and expression of each of the eight dancing men are unique in each figure. Some wear beards, some are clean shaven. They all wear the simple traditional Indian dress of a dhoti and an upper shawl called angavastram. The dhoti is worn draped and tucked between the legs, with folds hanging down center-front, creating a pant-like garment. This way of wearing a dhoti is called kachai. The upper cloth or shawl is worn around their neck, with the long ends hanging over the shoulders and down behind the back.

Fig. 40
Except for a few simple bracelets the eight men do not seem to wear any jewelry. We do not see the usual necklaces, belts, upper arm bands or anklets commonly worn by human figures from the medieval period. They could possibly wear ear studs or hangers, but this cannot be determined for sure from my photos. They all do display a very unusual hairdo or headdress reminding me of a Phrygian cap. A Phrygian cap is “a soft conical cap with the apex bent over, associated in antiquity with several peoples in Eastern Europe and Anatolia, including Phrygia, Dacia, and the Balkans”[10] (fig 40).  With a few of the figures it almost certainly looks like a cap. With others it could also be a kind of hair-top-knot, which is usually associated with rishi figures, except that here it also displays this forward bend typical of the Phrygian cap, which we never see with the rishis. Or these are perhaps caps with decoration, which sometimes was a feature of Phrygian caps. Some men display what may be a shawl or festoons or other decoration flying away from the headdress.

I have named the pillar of which the mālā element is largely intact as pillar I, and the other one, with the mālā element only partially intact, as pillar II. I have numbered the men as ‘a’ to ‘d’, in the clockwise direction, starting with the figure carrying the round object (see figs 38 and 39).

The Men of Pillar I

Fig. 41
I a: With his left leg crossed in front of his right this man is facing into the clock-wise direction. He wears a trimmed beard and moustache. His body is that of a healthy forty-something. The frills of his dhoti fall between and behind his legs. We see the folds of his upper cloth or shawl around his neck going over his shoulders. With his right arm he holds a round object on his shoulder. His left arm and hand crosses in front of his pelvis. An unusual, tassel-like ornament hangs from his left ear, or possibly belongs to his headdress, which can best be described as a Phrygian Cap.

I b: This man walks in clock-wise direction while looking backwards. His face is a bit damaged, but he seems to be clean-shaven. His left hand rests on his hip. His right arm is loosely hanging down by his side. His hair is separated in the center of his forehead, with a thick hair-do surrounding his face; or this element could be part of his Phrygian cap.

I c: The third man of this pillar faces out with his legs in what is known in South-Indian dance terms an arai-mandi, or half-sitting, position, with his legs bent and spread sideward. His right arm crosses down in front of his torso, while his left is bent with the elbow up and out near his left ear, while the hand is in front of his chest. His head is slightly bent toward his left shoulder. He wears a trimmed beard and moustache. His headdress has the same conical and forward bended form as with the other men, but this  could be hair tied up with sashes, like a turban. The way his dhoti is tied with the pleats hanging down is very clearly depicted. His upper cloth folds around his neck. He wears a bracelet on his right wrist.

I d: This figure we can see only from his left and his right side. His leg position shows him moving in a clock-wise direction. His is a flat, elongated face, with beard and moustache. His headdress is also conical and forward bending, with sashes and decorations tied around it. Some decoration hangs down the left side of his head. This could be ear jewelry or belong with the headdress. His left arm is hanging loosely down his side. His right arm is raised above his head.

The Men of Pillar II

Fig. 42
II a: This figure also holds a round object on his shoulder with his right arm. His left hand is on his left hip. His face has round features, is clean-shaven, with a flat nose, thin lips and round eyes. His hair is separated in the center, and otherwise lies flat against his skull. On top of his head is a forward bending ornament of some sort. From his head, or from his ear, some tassel-like feature hangs down. The position of his legs is that of arai-mandi, his legs spread outward with knees bent. His dhoti frills hang between his legs. His upper cloth drapes in front of his neck and over his shoulders. One end can be seen hanging down his waist.

II b: He has his legs open and bent in arai-mandi position. His face is elongated with a trimmed beard and moustache, a flat nose and full lips. His dhoti has the folds hanging between the legs. The shawl is draped across his neck. His left arm is loosely hanging down his side. His right arm is bent at the elbow with the hand at shoulder level. His headdress could be a top-knot decorated with tassels, or a pointed cap with decorations. Some decoration hangs down by his ears.

II c: This man moves in anti-clockwise direction with his left leg crossing in front of his right. At the same time he is turned backward and looks at the man on his right-hand side. Their hands, held in what could be a fist, or some ritual mudra or dance hasta gesture, touch. The folds of his dhoti flutter behind him. The shawl crosses in front of his throat. His is an elongated face with beard and moustache. His headdress is elaborately decorated with a band around his forehead and possibly other elements at the top and behind. There seems to be some object between him and the man on his right, but it is not clear what.

II d: This is the only figure who wears a sacred thread. It hangs over his left shoulder and drapes under his right arm. He does seem to have a moustache, but it is not clear from my photo whether he has a beard. He has round features. He walks towards his left, anti-clockwise, with legs one behind the other. His dhoti folds between his legs. His upper cloth drapes around his throat. His right arm is raised above his head. His left arm is bent at the elbow, the hand with folded fingers touching the hand of the man on his left side. His elaborate headdress or hair-do is built up above his head and is pointed and bent forward. It could be a cap with side flaps or tassels.

Water Step

The leg positions of these eight men suggest they could be performing a choreographed movement, a dance. As a dancer with extensive experience in both Bharata Natyam, Ballet and folk dancing, my analysis concludes they are moving in what is today internationally known as the Mayim, the Water step. This is one of the basic steps from folk dancing, where the footwork creates a waving motion, reminiscent of the waves of water. The Western legend of this step is that it was created by Jewish settlers in a kibbutz in the desert to commemorate the finding of water after a seven-year search.[11] But in a television report on a kumbhabhishekam many years ago, I saw a group of Shaiva Pandits burst into a spontaneous dance performing exactly this Water Step.[12] Offering the understanding this very basic and beautiful foot movement is also part of the Indian Vedic and Agamic tradition. The foot movements consist of a cross-in-front, step aside, cross-behind, step aside. It can be executed in clockwise and anticlockwise direction. When taken together we could reconstruct the footwork depicted on the two pillars to represent this Water Step.


These two pillar fragments on the one hand fit perfectly within the architectural and artistic idiom of medieval South India. The presence of a base with inverted and upright lotus petal moldings, and the festoons of the robust mālāsthāna place them solidly within the framework of South Indian temple art.

On the other hand, several characteristics place them outside any of the known categories. The pillar fragments are round all through from base to mālā. Both the base and the mālāsthāna are decorated in a style which lies outside the known idioms. The two groups of four Dancing Men circling the shaft are unique both in subject matter and in style. The men represent various social backgrounds, one wearing a sacred thread, the other seven not. Their facial features may even indicate they represent people of diverse ethnic background, from within or without the Indian subcontinent. Although they all wear the traditional Indian dress, their Phrygian caps or hairstyle stands out against everything we know so far.

As far as placing them within the context of South Indian chronology, my conclusion must be they definitely do not fit within the Chola, Later Pandya, Vijayanagara or Nayaka categories. Neither can we establish a stylistic link to the competitors of the Pallavas, the Chalukya and Rashtrakuta dynasties. The robust and bold mālāsthāna may be an argument for a Pallava provenance. Possibly around the period of the building of the Kailāsanatha as that temple shows a mālāsthāna which to some extent similar, although not of the same vigor. Their presence in the very center of Kanchipuram is also an argument for a Pallava provenance.  But the element of the Dancing Men counters such a conclusion. There is nothing in all of the known Pallava period which even hints at this unique feature.

The inevitable conclusion is that although these pillars belong solidly within the South Indian artistic idiom, they are unique and an anomaly and may represent the expression of a so far unknown phase or dynasty. Possibly earlier or contemporaneous with the Pallava dynasty, they remain an anomaly in Pallava land.

Liesbeth Pankaja Bennink combined her history study at the Rijksuniversiteit Utrecht (The Netherlands) with her training as Bharata Natyam dancer under Smt. Rajamani Knols (Amsterdam, Mumbai, Bangalore). After accomplishing her Arangetram in 1981 and her graduation in 1983 she performed and taught for many years in The Netherlands and abroad. Seeking a deeper understanding and knowledge of Indian dance in the context of the ancient traditions in which it is rooted, she entered into a long-time collaboration with Raja Deekshithar (R.N.Natarajarathina Deekshithar M.A.1949-2010). He was an independent researcher and a scholar belonging to the Podu Deekshithar priestly community of the Chidambaram Shri Shiva Nataraja temple. The past years she has pursued her research interests in Indian history, arts, and traditions. Some of her subjects relate to Bharata Natyam dance and its evolution, others to astro-archaeology and the astronomical meaning hidden in mythology and symbolism. She is also studying the meaning behind the symbolism of temple traditions and mythology. As an enthusiast of Indian history, Liesbeth Pankaja has been translating excerpts from the writings of Jacob Haafner, an 18th-century Dutch traveler and author on South India, from Dutch into English. This material is relevant to the understanding of India in the last part of the 18th century but was so far largely unavailable in English. She is looking forward to sharing her insights with all those interested in Indian civilization and its history. For further reading, you can visit her website and blog. Liesbeth Pankaja can be reached at Liesbethpankaja[at]


1. It is of course possible there is an oral tradition. But I didn’t ask anyone at the time.

2. It would be possible to calculate their original height using the ratios of length and diameter given in the Shilpa Shastras, architectural texts like the Manasara and Mayamata.

3. Jouveau-Dubreuil, G., The Pallavas, Pondicherry 1917; Gopalan, R, History of the Pallavas of Kanchi, Madras 1928; Heras, H, Studies in Pallava History, Madras 1933; Lockwood, Michael, Pallava Art, Madras 2001.

4. Hoekveld-Meijer, G, Koyils in the Colamandalam, Typology and Development of Early Cola Temples, Amsterdam 1981, p. 69, 246-272; Meenal Kumar, R., Pillars [Stambha] - the supportive elements of Hindu temples, International Journal of Current Research Vol. 9, Issue, 05, pp.50101-50107, May, 2017, Available online at ; Ram Raz, Essay on the Architecture of the Hindus, London 1834, p. 28-40.

5. See figures 20 and 21.

6. A pilaster is an ornamental element on a wall, looking like a pillar but without a structural or supportive function. In Indian temple architecture it is used to create and underline symmetry in lay-out and ground plan.

7. Sivaram, Rama S, Origin and emergence of style in early Chola art, Volume 1, Baroda 1989

8. A yali is a mythological being usually depicted as a horned lion. Often depicted in Indian temples.

9. Lockwood 2002, p 109-115.



12. Many years ago, while staying with my Deekshithar friend Raja Deekshithar and his family in Chidambaram, India, I watched a short report about a kumbhābhiśeka on the television. I don’t remember which temple was the subject, or which network broadcasted it. There were images of the yajña and the procession taking the vessels with consecrated water to the vimāna via the prākāra. At one point suddenly a small group of Iyars, beautifully and elegantly dressed with kachai dhoti and aṅgavastram, neatly tonsured and bejeweled, burst into a dance. They formed a circle and performed a short group dance, circling clock-wise. As a dancer I recognized the steps they performed as a step well known in Western folk dancing as the ‘mayim step’. The Western legend of this step is that it was created by Jewish settlers in a kibbutz in the desert when they found water after a seven-year search. The footwork of this step generates a waving movement, in imitation of water waves. I was thrilled to see this step performed as part of the kumbhābhiśeka rituals. It looked like a spontaneous bursting out in dance, from the bliss experienced in the moment. I realized this group dance must have a far greater age and deeper meaning within the Indian traditions, besides the understanding in Israeli folklore.

Download the PDF version of this article | articles