Chapter 7 - The age of the Manasara

IN TRYING TO establish, in the preceding chapter, a relation of influence between Vitruvius and the Manasara, one of my objects was to find out a landmark for the latter, as the date of the former is known approximately. The * From the fragments of inscriptions relative to the Vitruvia family found in the neighbourhood of Farmiae (the present Mola di Gacta), is has been presumed, without a great stretch of probability, that it was in this territory Vitruvius was born. The age in which he lived. was doubtless between the time of the death of Julius Casar and the battle of Actium, though some have assigned it to the reign of Titus. But his omission of the mention of a great number of magnificent buildings, erected after the time of Augustus and his especial mention of the theatre of Pompey as the only one of stone, sufficiently prove that such a conjecture is not warranted by circumstances. dedication, moreover, points to Augustus as the patron [of Vitruvius] and the incident of C. Julius, the son of Masanissa, who was born'in the army of Julius Casar, having lodged with him, as related in the third chapter of his eighth book, seems clearly to indicate the time of his existence. It is likely that the following treatise (of Vitruvius) was composed when he was advanced in life, and that it was presented to his patron after he had assumed the title of Augustus, that is, twenty-five years before the Christian era, inasmuch as he speaks of a temple erected to Augustus, in his Basilica at Fano."¹ < The other landmark may perhaps be supplied by the following thesis admitted by Leiden University. There seems to have been a relation of indebtedness between the Manasara, on the one hand, and on the other hand, the architectural portions of the Agni-Purana, The Garuda-Purana, the Matsya-Purana, and the Bhavishya-Purana, the Kamikagama, the Suprabhedagama, and the Brihat-samhita.' 'The reasons for and the arguments in favour of such a belief have been discussed in great detail elsewhere, 2 and need not be repeated here. For further scrutiny and more minute comparison, the Brihat-samhita 1 Professor Gwilt, Preface, xii. See pp. 186-209. 240

of Varahamihira, one of the nine gems¹ at the court of a mythical Vikramaditya, is selected. Professor Kern has given a probable date, A. D. 550, to the Brhat-samhita. The ages of the Puranas and the Agamas mentioned in the thesis are more conjectural. Although primarily not a treatise on architecture, the Brhatsamhita has devoted five chapters² (LIII, LVI, LVII, LVIII and LXXIX) to this art. Three of these, called Vastu-vidya or the science of architecture, Prasada-lakshana or the description of temples under twenty types, and Sayyasana-lakshana or the description of bedsteads and couches, deal with architecture proper, and the other two, called Vajralepa or the first casting of image, and Pratima-lakshana or the description of images, briefly refer to sculpture. The following similarities between the Manasara and Brhat-samhita may be noticed : Brihat-samhita Origin and development of the science of architecture Objects of architecManasara I, I ' } LIII, 1. ture. III Testing of oil IV-V LIII, 2-3 (details differ). LIII, 95-97, 85-92, 115-117. Site plan VII LIII, 42-69, 83-84. Offerings to deities VIII LIII, 99-100. Dimension of storeys CHI LIII, 4-26, LVI, 29-30. Columns XV Temple-buildings XVIII LIII, 27-30, 112-113, 121-123. LVI, 3-8 (site), 9 (ground), 10-16 (general), 17-19 (twenty types). One-storeyed buildXIX LVI, 23, 26. Five-storeyed buildings .. XXIII LVI, 27. 1 dhanvantarih ksapana kamarasimha kuveta labha ghataka rikalidasah | khyato varahamihiro nrpateh sabhayam ratnani vai vararucirnava vikramasya || _ - ( jyotirvidabharana, Kern, B. S. 17 ). The existence of these gems as contemporaries has been held to be untenable. This refers to Kern's edition published by Royal Asiatic Society, London. In some other editions, for instance, in that of Sudhakara Dvivedi, Benares, Vikrama era 1658, these chapters have got a different numbering. Our references to the Btihatsamhita are mostly to Kern's edition. 241 p

Manasara Brhat-samhita Six-storeyed buildings .. Seven-storeyed buildings Eight-storeyed buildings XXIV LVI, 22. XXV LVI, 24. XXVI LVI, 21. Ten-storeyed buildings .. XXVIII LVI, 20. Halls and pavilions XXXIII, LIII, 118. XXXIV Situation and meahouses XXXVI LIII, 70. Ceremonial surements of entry into a new y-built house Situation and measurement of doors Phalli Images of female deities Images in general Largest type of tentala measures First images LIII. 26-27, 70-82 LVI, 10, 12-16. XXXVII | LIII, 125. XXXVIII, XXXIX LII LVIII, 53-55. LIV LVIII, 56. LXIV LVIII, 31-52, (ends • abruptly). casting of LXV LVIII, 4-30. LXVIII LVIII, 1-8. An elaborate discussion has already been introduced regarding the types of buildings and the five orders.2 Two other points of special and general importance may be elaborated here. The site-plans are treated under twenty-four schemes in the Manasara. Of these, descriptions in detail are given of the eighth and the ninth schemes, which consist respectively of sixty-four and eighty-one squares. In the Manasara it is stated, by way of explanation, that these two plans were much in use. Varahamihira also has described only these two plans. In the Brihat-samhita there is not ² See pp. 201-206. ¹ See pp. 187-197. 242

the slightest reference to the other twenty-two schemes. As is usual with him, Varahamihira has changed the names or location of the squares here and there. Another striking point of similarity is that only the square plans are described in both the treatises, although in the Manasara five shapes or forms are given to the buildings. Varahamihira also has referred to the round type of buildings. Corresponding to these shapes there should be the ground plans also. But these are unexpectedly missing in the Manasara and also in the Brhatsamhita. The details of round or circular plans and also of triangular plans, both consisting of eighty-one squares, have been quoted from a mythical Bharata-muni by Bhattotpala, a commentator of the Brhatsamhita. So, in matter of such a striking omission also, Varahamihira seems to have faithfully followed the Manasara, Garga, or whatever else his sources might have been. The other point proposed to be discussed here is that concerning the sources of the Brhat-samhita in architectural matters. Varahamihira says that the science of architecture has come down to him from Brahma (Kamalabhu) through several generations of sages. He further admits that all matters relating to architecture are taken from Garga, and small portions of the architectural treatises of Manu and others have been put in from memory.¹ The names of the sages passed over here may be gathered together from casual references. Mention is made of Vasishtha, Maya, Visvakarman, Bhaskara and Nagnajit." The Puranas are not mentioned by Varahamihira. But some of the Puranas are no doubt carlier than the Brihat-samhita.³ 3 1 prasadalaksanamidam kathitam samasad | gargena yadviracitam tadihasti sarvam || manvadibhirviracitani prthuni yani | adeqla ala gara gasferre: 11- (Brhat-samhita, LVI, 30-31). caturangulam vasisthah kathayati netrantakarnayavivaram | -LV 111, 8. mathakathito yogoyam vijneyo vajrasanghatah | -- LVII, 8. bhumikangulamanena mayasyastottaram satam | sardhahastatrayam caiva kathitam visvakarmana | LVII, 29. sarvapratimasvevam subhasubham bhaskaroktamamam | LVIII, 52. asyam sakesanicayam sodasaderyena nagnajit | LVIII, 15. nagnajita tu caturdasadena dravidam kathitam | -VIII, 4. See pp. 277-278. 243

It has been shown that with regard to the techincal names and other details of the twenty types under which temple-buildings are described, the Matsya-Puranna ( Chapter CCLXIX, vv. 28-53), the Bhavishya - Purana (Chapter CXXX, vv. 2735 ), and the Brihat -samhita (Chapter LVI, vv. 20-28) are identical. 1 The Bhavishya - Purana ( Chapter CXXX, vv. 15-26, 36 and 37, 27-35) can be read, letter for letter, in the Brihatsamhita (LVI, 8-19, 30, 20-28). When verses 22, 36 and 37 of the former are compared with the identical verses 15, 29 and 30 of the latter, it seems as if Varahamihira were the debtor. It should be noted that the linguistic defects of the Bhavishya-Purana are removed in the Brihat-samhita. Similar illustrations can be drawn from the Matsya-Purana and the Brihat -samhita also. For instance, 3 of the former, verses 2 ( Chapter CCLV), 19 and 20 (chapter CCLXX) can be compared with the latter in respect of LIII, 28, and LVI, 12, 13. Varahamihira's is apparently the improved version in the revised edition. On this ground one is tempted to place these Puranas before the Brihat-samhita. But 1 See pp. 190-195. 3 2 sailamanagalyavihagah srivrksah svastikairghataih | manastamena bhagena pratima syatsapindika || (22). sahasratritayam caiva kathitam visvakarmana || (36). prahuh sthapatayascatra matamekam vipascitah | kapotapaliniyuktamato gacchati tulyatam || ( 37 ) . and sesam mangalyavihagaih srivrksah svastikaighamteh | mithunah patrapallibhih pramathaiscopasobhayet || (15). sarddham hastatrayam caiva kathitam visvakarmana || (29). prahuh sthapatayascatra matamekam vipascitah | kapotapalisamyukta nyuna gacchanti tulyatam || ( 30 ). rucakaracaturahsyattvastastro vajrocyate | ( 255, 2). and samacaturastro rucako vajro'stastridvivajra ki dvigunah | - LIII, 28. vistarardho bhavedgarbho bhittayo'nyah samantatah | garbhapadana vistirnam dvaram trigunayatam || ( 270, 19). and vistarartho bhavegdarbho bhittayojyah samantatah | garbhapadena vistornam dvaram dvigunamucchritam ||LVI, 12. again tatha dvigunavistirnamukhastadvadudambarah | vistarapadapratimam bahulyam sakhayoh smrtam || ( 270, 20 ) . and ucchrayatpadavistirna sakha tadvadudambarah | vistarapadapratimam bahulyam sakhayoh smrtam || -LVI. 13. 244

arahamihira himself has not admitted his debt to these authorities. In these circumstances priority might be claimed for him. There is one other point which deserves special notice. In the Matsya-Purana eighteen professors¹ of the science of architecture are mentioned, namely, Bhrigu, Atri, Vasishtha, Visvakarman, Maya, Narada, Nagnajit, Visalaksha, Purandara, Brahman, Kumara, Nandisa (Siva), Saunaka, Garga, Vasudeva, Aniruddha, Sukra, and Vrhaspati. Of these eighteen professors, Garga, Maya, Visvakarman, Vasishtha and Nagnajit have also been mentioned in the Brihat-samhita. Varahamihira, the author of the Brihat-samhita, has included Bhaskara and Manu, who are not met with in the MatsyaPurana. The identity of these mythical sages is a vexatious matter in Sanskrit literature. One Bhaskara or Bhaskaracharya was the author of the Lilavati and the Siddhanta-siromani. Of Manu we shall presently speak more. But the Matsya-Purana does not include these names in its lists, nor does it mention Varahamihira. In the Manasara there is a list of thirty-two authorities, namely, Visvakarman, Visvesa, Visva-sara, Prabodhaka, Vrta, Maya, Tvashtar, Manu, Nala, Mana-vid, Mana-kalpa, Mana-sara, Manabodha, Prashtar, Visva-bodha, Naya, Adisara, Visala, Visva-kasyapa, Vastu-bodha, Mahatantra, Vastu-vidyapati, Parasariyaka, Kala-yupa, 1 Matsya-Purana, Chap. CCLV, 2-4. Compare: iti proktam vastusastram purva garbhaya dhimate | gargatparasarah praptastammatprapto brhadrathah || brhadrathadvisvakarma praptavan vastusastrakam | sa eva visvakarma jagato hinayakathayatpunah || vasudevadisu punarbha lokam bhaktito'bravit || ·(Visva-karma-prakasa, Benares, 1888, XIII, 25 to 27). proktani pamcaratrani saptaratrani vam maya || vyavastani munibhirloke parcavisati samkhyaya | hayasirsam tantramadyam tantram trailokyamohanam || vaibhavam pauskaram tantram prahaladam gargyagalavam | naradiyam ca samprasnam sandilyam vaisvakam tatha || satyoktam saunakam tantram vasistham jnanasagaram | svayambhuvam kapilam ca tarksya narayaniyakam | atreyam narasimhakhyamanandakhyam tatharunakam | baudhayanam tatharse tu visvoktam tasya saratah || - (Agni-Purana, Chap. XXXIX, v. 1-5). 245

Chaitya, Chitraka, Avarya, Sadhakasara-samhita, Bhanu, Indra, Lokajna, and Saura. In the opening verse it is stated that the science of architecture has come down to the sage Manasara from Siva, Brahma, and Vishnu, through Indra, Brihaspati, Narada, and all other sages. In a mythical genealogy of the artists it is further stated that from the four faces of Brahma originated the four heavenly architects, namely, Visvakarman, Maya, Tvashtar, and Manu. Their four sons, called respectively, Sthapati or the chief architect, Sutra-grahin or the designer, Vardhaki or the painter, and Takshaka or the carpenter, represent the guild of the modern architects." It should be noted that Visvakarman, Maya, Manu, and Tvashtar 2 are mentioned twice, once to represent the heavenly architects, and again as modern architects. In the same sense Indra is also mentioned twice. Visvakarman and Maya, to whom many extant architectural treatises are ascribed, are common in the Manasara, the MatsyaPurana, and the Brhat-samhita. The Manasara and the Matsya-Purana have, therefore, in common five authorities, namely, Brihaspati, Indra (under the name Purandara in the Purana), Visalaksha (alias Visala in the Manasara), Visvakarman, and Maya. The Manasara and the Brhat-samhita have in common Visvakarman, Maya, and Manu. Visvakarman, etymologically implying the Creator of the universe, is more or less a professional name for an architect. Manu is less so. This is a generic name. Mention is made of fourteen Manus,3 namely, Svayambhuva, Svarochisha, Auttami, Tamasa, Raivata, Chakshusha, Vaivasvata, Savarni, Daksha-savarni, Brahma-savarni, Dharma-savarni, Rudra-savarni, Rauchya-daiva-savarni, and Indrasavarni. Manu is a sort of second Creator, the Indian Adam, representative of man and father of human race. It seems clear, however, that there must have been an architect Manu also in the ordinary sense of the term, because with him several architectural works are associated. He is stated in the Ramayana to have built the ancient city of Ayodhya, the capital of king Rama. 1 Chapters LXVIII, I, II, see pp. 155, note 3; 97, 98. 2 Tvashtar is mentioned in Padmagupta's Navasahasanka-charita, as the sculptor of a crystal sivalinga (phallus). ³ Manusamhita, I, 63. 4 ayodhya nama nagari tatrasillokavisruta | manuna mazda a ge fafa Fagy 11-(Ramayana, Adikanda, verse 6). 246

aya is a more historical person. Several existing architectural treatises are ascribed to him.¹ He may not be as old as the ZendAvesta. Ahura-Mazda and Maya-Asura are perhaps not one and the same person. But he is mentioned in unmistakable terms as the architect of a wonderful council hall, of which it is stated there could not be any parallel in the world of the mortals, and whereon all heavenly ideas were depicted in bricks and stones. He declares himself as a great poet of architecture (mahakavi), a Ruskin, among the rivals of gods, and he is their Visvakarman, who was the heavenly architect among the gods.2 The town of Ratnavati is stated to have been built by Maya, the prince of demons, in Padmagupta's Navasahasanka-charita. Like Manu, Maya is also a generic name. He is also known by some other personal names.³ So the Maya of the Manasara, of the Matsya-Purana, and of the Brihat-samhita may not be one and the same person. It is just possible that there might have been a Maya who borrowed from, or based his treatises in any case, upon the Manasara.ª In fact it is perfectly clear from the list of authorities quoted from the Manasara that there must have been at least one more Manasara, from whom or from which our Manasara has borrowed. It has also 1 See the writer's Encyclopaedia, Appendix I, where a note on the latest discoveries of the Maya civilization in America is also given. 2 aham hi visvakarma vai danavanam mahakavih | so'ham vai tvatkrte kincitkartumicchami pandavah || (5) tato vicintya manasa lokanathah prajapatih | codayamasa tam krsnah sabha vai kriyatamiti || (9) yadi tvam kartukamo'si priyam silpavatam vara | dharmarajasya daiteya yadrsi miha manyase || ( 10 ) yam krtam nanukurvanti manavah preksyadhisthitah | manusyaloke sakale tadrsim kuru vai sabham || ( 11 ) yatra divyanabhiprayan pasyema hi krtamstvaya amuranmanupascaiva sabham tam kuru vai maya || ( 12 ) - -(Mahabharata, Sabha-parvan, Chap. I, 5, 9-12). The famous commentator Nilakantha adds the following note: visvam karmakrti sadhyam yasya sa visvakarma | mahakavih silpapanditah | asuran manusan ityupalaksanam devagandharvadinamapi abhiprayan lepacitre lekhyacitram ca caturdasabhuvanantarasthatattajjatiya svabhavikananavidhalilapradarsanena manovrttih pasyema yaddarsanena brahmandantaravati sarvam vastujatam drstaprayam bhavatityarthah | 3 See pp. 159-161. 4 See P. 16. 247

been pointed out that the term "Manasara " has been used to imply both a person and a treatise. The uncertain identities and the confusing chronology are indeed stumbling blocks in the field of Sanskrit researches. In all items of comparison between the Manasara and the cognate works, we have seen² that the Manasara contains fuller lists. In the present instance also there are more than thirty-two authorities mentioned in the Manasara, while the Matsya-Purana is content with a list of eighteen, and the Brihat-samhita has specified only seven. But none of these three treatises has admitted the authority of either of the other two. From this it would appear as if they were quite ignorant of the existence of one another, being separated by an insuperable gap in time or space. Such a relationship is untenable, I should say improbable, between the Matsya-Purana, the BhavishyaPurana, and the Brhat-samhita in any case, unless, however, we choose to suppose that there might have been an unknown authority or some floating tradition by which these treatises have been influenced in the same way, even to the extent of chapter and verse, but without any knowledge of one another. I have failed to satisfy myself with such a hypothesis. For we have seen identical passages in these works. All these three contain the same list of twenty types of buildings, bearing the same technical names and identical in other details.4 Buildings are described under certain types in all the architectural works. Their technical names have no signification. Unless one list is copied from the other these names need not be identical. In fact such is the case with regard to the fuller list in the Manasara. Therein we have seen ninety-eight types of buildings described under more architectural divisions and with fuller achitectural details than in these non-architectural works. Except in one or two solitary instances like Kailasa, the names of these types of buildings are not identical. There are certain similarities. For instance, the merukanta of the Manasara is read imply as meru in the Puranas and the Brihat-samhita. This is certainly an improved reading, first, because meru as the name of a mountain or as a geographical term is well known in Sanskrit 3 1 See Preface, First edition (Indian Architecture, p. 3). 2 See pp. 187-208. 4 See pp. 191-195. 248 3 See pp. 245-247. 5 See pp. 187-189.

literature, and secondly, kanta, in the expression merukanta, is meaningless. Similarly, the reading vrtta of the Puranas and the Brhat-samhita is an improved version, a good amended form of vivrta of the Manasara. Almost similar is the case with regard to another architecturally very important object, namely, the column or order, for the columns are stated by the authorities to be the regulator of the whole composition. In this case also the Manasara contains a fuller list. It has two sets of five technical names for the orders, while the Matsya-Purana and the Brhat-samhita contain only one set of the five orders. The names of these orders in the Manasara are different from those in the Purana and the Samhita, but Varahamihira has given the very same five names to the orders as the Matsya-Purana, and they have also the very same eight names for the mouldings or the component parts of a column. The Manasara, as in all other cases, has a fuller list of mouldings also. It contains more than forty-seven mouldings for the pedestal, base, shaft, and entablature, the shaft being given five special mouldings. And as in the case of the types of buildings, there are some names of mouldings, for instance, ghata and hara, common in the Manasara, the Matsya-Purana, and the Brihat-samhita. So in three important architectural matters, namely, the preceding authorities, the types of buildings, and the orders and their compoment parts, the Manasara has fuller lists than those in the MatsyaPurana and the Brhat-samhita, which are exactly identical in these matters. In these circumstances one is ordinarily likely to think that a later work only can make a thing more complete. But there is another essential point to consider, namely, that the Manasara is avowedly an architectural treatise, while the Matsya-Puranna and the Brihat-samhita are not. Their treatment of architectural matters is but casual, and in fact they have entirely left out purely architectural description. It is clear beyond doubt that the Purana and the Samhita must have consulted an architectural treatise for their information and guidance in achitectural matters, just as they have, certainly, based their references, for instance, on medicine, to a standard medical treatise. If the Manasara had an opportunity of consulting Varahamihira or the Matsya-Purana, the reading like vivrta for vrtta, or 1 For instance, Naishadhacharita, 16, Bhartrihari, Vairagya-Sataka, 150, etc. Compare the terms like Sumeru, Uttarameru, etc. 2 See pp. 201-206. 249

merukanta for meru, could not have remained unamended in it. Besides, if the Manasara had been composed after the works like the Matsya-Purana and the Brihat-samhita, why should it not have added. these two to its long list of authorities? It would be no argument to say that the author of the Manasara may not have consulted these authorities or may have been quite ignorant of their existence. For, though not primarily works on architecture, the Matsya-Purana and the Brihat-samhita have been well known to subsequent literature, and we shall presently show that the author of the Manasara had an extensive knowledge of things from a wide study and observation. In these circumstances, though ready to re-adjust my views in the light of new facts, my present impression is that there must have been a direct influence between the Matsya-Purana, the Brhat-samhita, and the Bhavishya-Purana; while the connexion between these treatises and the Manasara may have been indirect. The age of the Manasara, however, is indicated by other things also, and these will be discussed presently. There are only a few treatises wherein the term "Manasara "is mentioned. The Agni-Purana, as already pointed out,1 has some passages of uncertain meanings wherein the term occurs. For instance, it is stated that above the suka-nasa (lit. parrot's nose) or gargoyles, that is, the water-spout in a building, there should be a vedi or platform furnished with a neck. And this should be as prescribed in the Manasara (manasaraka), or the object of it is to make a passage for refuse (malasaraka). This latter interpretation seems untenable, for the adjective is used in the neuter singular and ordinarily would not qualify a feminine singular noun. If the first rendering be acceptable, the expression would form a separate clause, iti manasarakam, meaning this is in accordance with the rules of the Manasara. 2 There are reasons to think that a relation of direct influence exists between the Agni-Purana and the Garuda-Purana.3 And, through the latter, the former may be connected with the Matsya-Purana, the Bhavishya-Purana, and the Brihat-samhita." The Sukra-niti is another important work, which, though not an architectural treatise, deals largely with subjects relating to 1 See Preface, First edition. Compare also p. 245, note 1. tadurdhvam tu bhavedvedi sakantha manamarakam ? ( manasarakam or malasarakam ) | 3 See pp. 189-191. -(Agni-Purana, XLII, 17). 4 See pp. 191-195, and the writer's Encyclopaedia, under Prasada. 250

architecture and sculpture.' It is a work on royal polity ascribed to an author, Sukracharya, whose age has not been clearly established. It appears to be anterior to the Matsya-Purana for the reason that the latter has included Sukra as one of its eighteen authorities.2 The question of the identity of Sukracharya with this Sukra must necessarily come in. But there is hardly a satisfactory answer to give. In the Sukra-niti we notice, also, a large number of passages common to it with another work called the Kamandakiya-niti. This has been assigned by Dr. R. L. Mitra to the fourth century of the Christian era on the ground of its dedication to Chandra Gupta, existence of Hindu temples, and the absence of any trace of Buddhism in the fifth century A. D. when the Chinese traveller Fa-hien visited Java, where, in the island of Bali, the work was discovered.³ This Kamandakiya-niti, which has apparently borrowed from the Artha-sastra of Vishnu Gupta, seems, in its turn, to have been freely drawn upon by the Agni-Purana.5 This Purana, we have shown, may have borrowed from the Manasara also." This introduction of the Sukra-niti, the Kamandakiya-niti, and the Artha-sastra, together with the Matsya-Purana, the Agni-Purana, and the Manasara, may appear as an episode. But a time may come when the inter-relation of all these treatises will be more satisfactorily established. 1 The next external references to the name of Manasara are met Chapter IV, Section 4 (1) devamandiradinirmanavyavastha ( 2 ) pratimanirmanavyavastha etc. (see details in Appendix I, in the writer's Encyclopaedia). Page 164. 3 yasya prabhavadbhuvanam sasvate pathi tisthati | ajahara nrcandraya candraguptaya medinim || (Kamandakiya niti, 1-5). Here, it is argued, Chandra Gupta refers to the first or second Chandra Gupta of the Imperial Gupta dynasty, who are assigned respectively to A. D. 320-326 and a. D. 375-413. 4 Alfamearga NaraSMETAZITI: | samuddadhe namastasmai visnuguptaya vedhase || - (Kamandakiya, L. 6). Dr. Jacob places the Artha-sastra in the fourth century B. C. (Berlin Academy Seitzungaberichte, 1911, pages 954-973, 1912, pp. 832-849). Professor Keith tends to bring it down to the second or first century B. C. (J. R. A. S., 1915). Dr. R. L. Mitra, Kamandakiya, Bibl. Ind., p. 4. G See pp. 110-118. 251

with in a famous prose romance, the Dasa-Kumara-Charita,¹ by a very eminent author, Dandin, who probably dates from the sixth century A. D.' Therein Manasara is repeatedly mentioned in unmistakable terms as the king of Malava (modern Malwa), with whom was engaged in war king Rajahamsa of Magadha or Pataliputra, the modern Patna. The latter was the father of Rajavahana, the chief of the ten princes or Dasa-Kumara, after whom the work is named. Here is a possibility of the Manasara being connected with the king of Malwa bearing the name Manasara. There are several works in the Sanskrit literature which seem to have been named after their patron, for instance, the Skanda-Purana is supposed by some scholars to have been associated with the name of Skanda-Gupta of the Imperial Gupta dynasty. The Harsha-charita has undoubtedly been named after king Harsha. But nothing more is known about king Manasara of Malwa, nor is anything stated, directly or indirectly, about him in the Manasara itself. On the other hand, the internal references to the expression Manasara, which have been already introduced elsewhere," prove that the term has been used in three different senses, namely, a treatise, an architect, and a class of sages or professors of architecture bearing the surname, like Manu or Maya, or the professional epithet, Manasara. In none of these senses, however, would the king of Malwa fit in. If he were a real personage and had any connexion with this standard treatise on architecture, and preferred to remain incognito, the author of the Manasara would have added a fourth ambiguity referring to his anonymous patron. In the body of the Manasara there are several passages, which will be presently discussed, evincing on the part of its author not only a clear knowledge of man and things of the then Magadha and Malwa, but also of all other chief cities and the broadest divisions of India of his time. 1 Edited by Kale, Bombay, 1917; see page 4, repetition below is felt unavoidably necessary: (i) Page 4, para. 2, line 3-aarasi qizadi ga¶A¶rugene 4979 tamanasaram manasaram prati ... samgramabhilasena rosena mahatavisto niryayau | (ii) Page 8, para. I line 8 - malavanatha jayalaksmisanatho magadharajyam prajyam samakramya puspapuramadhya tisthat | (iii) Pages 12-13 - rajahamgo munimabhasata, bhagavan, manasarah prabalena daivabalena mam nirjitya madbhagyam rajyamanubhavati | • • (iv) Pages 13-14- avantisundari nama manasaranandini nagaropantaramyodyane viharo- tkanthaya manobhavamarcayanti reme | 2 See Preface, Indian Architecture, First edition. 252

The third external reference to (the architect) Manasara is found in two epigraphical records of uncertain reading. In these unpublished documents the epigraphist reads the expression, which is used in two inscriptions to imply the name of an architect, as Manasarpa and not Manasara. In the light of information presented for the first time in our volumes, the epigraphist may perhaps be ready to revise his reading of the expression when the inscriptions are properly edited and finally published.¹ In an architectural compilation, Silpa-samgraha, of apparently a very late date, we have shown already 2 that the Manasara is quoted in its true form. About the worth of considering this reference I am rather doubtful. Not that I am unwilling to bring down the Manasara, but because there are several facts which cannot fit in to a very late date like A. D. 1830, when a manuscript³ of the Manasara was copied. The compiler could have easily consulted the Manasara, even if the latter were placed side by side with Vitruvius, or before Maya-Asura of the Mahabharata, or Ahura-Mazda of the Zend-Avesta. Of the internal evidences from the Manasara the following points may be considered. For the orientation of buildings it was necessary for the Indian architects to ascertain the right cardinal points. For this purpose the Manasara, in agreement with all complete works on architecture, including Vitruvius, makes use of a gnomon, 4 obviously because the mechanism of the compass was not known to the ancient architects. For similar purpose the astronomical treatises also, like the Suryasiddhanta and the Lilavati and the Siddhanta-siromani of Bhaskaracharya, use the gnomon. The calculation of the shadow is the main object in this matter, and the gnomon is used simply to ascertain the shadow. The sun's rays falling on an object like the gnomon causes the shadow. So at first the obstructed light, which gives rise to the 1 Epigraphist's Report, Madras, 1901, nos. 207, 209. See page 4, note 2; 130, note 5; 176. 2 See page 106. 3 Called 'C' in the description of manuscripts attached to our edition of the Manasara. Bis dated 1677 of the Salivahana era (1823); 'D' is dated 1656 of the Saka cra (1734). The remaining eight copies-A, E, F, G, H, I, J, K, are not dated. 4 For full description, see the writer's Encyclopaedia, under Sanku. 5 Surya-siddhanta, III, 1-4, Lilavati, part 2, section 4, Chap. II; SiddhantaSiromani, last part, Chap. VII, 36-49. Cf. Vitruvius, Book I, Chap. VI; Book IX, Chaps. IV, VIII. The actual process of working the gnomon for ascertaining the cardinal points and dialling is described in full details from all these authorities in the writer's Encyclopaedia, under Sanku. 253

AGE OF THE manasAra HINDU ARCHITECTURE IN shadow, must naturally be taken from the sun. But the sun's light is uncertain and cannot be adjusted according to the requirements of scientific and advanced study. It is, therefore, not difficult to believe that the later astronomers easily found some artificial and adjustable light to replace the natural and unadjustable light from the sun. In the Manasara only the sun's light is made use of, while in the Surya-siddhanta and other astronomical works lamp-light was used in order to measure the shadow. These latter works also followed an improved method in ascertaining the level whereupon the gnomon was erected in order to calculate the movement of the shadow accurately. The methods followed in the Manasara are antiquated. It seems, therefore, that the Surya-siddhanta and other astronomical works must come after the Manasara. The next internal evidence presented here for consideration is that concerning the knowledge evinced in the Manasara of the most prosperous countries throughout India. We have seen that ninety-eight types of buildings are described in the Manasara under twelve divisions, namely, of one to twelve storeys. Although sixteen-storeyed or even seventeen-storeyed gate-houses (gopuras) are mentioned, religious or residential buildings are not erected beyond twelve storeys.1 The technical names of buildings of one to eleven storeys are more or less poetical. But the buildings of twelve storeys, largest and most gorgeous of all edifices, bear more significant names. They are called Magadha, Janaka, Madhya-kanta, Vamsaka, Virala, Panchala, Sphu (Gu)rjaka, Kerala, Dravida, and Kalinga. These are the names of ten countries which cover the length and breadth of India. At one time or another they seem to have been very prosperous, possessing as they did distinctive types of the largest and most magnificent edifices. Magadha is the country of South Bihar, where the Pali language was spoken. Janaka, or the country of some twenty-one generations of Janaka kings, otherwise called Vidcha, with capital city Mithila, is North Bihar, which corresponds to the modern Tirhut and Purniya divisions, between the Gandaki and Kosi rivers. Madhya-kanta 1 See pp. 52, 47-51, 111-113. 2 Magadha is also mentioned in the Dasakumara-charita, 3/4¶¶¶ goga which was conquered by king Manasara of Malava (see Preface, Indian Architecture, 1 st Ed., note). But Malwais not honoured with a separate type of twelve-storeyed buildings, and it would appear strange and unusual if this king Manasara were the patron of the Manasara, the standard work on architecture. 254

stands for the Madhya-desa, the middle country, or the tract situated between the Himalayas and the Vindhya range to the cast of Vinasana and to the west of Prayaga or Allahabad.¹ Some authorities make it the Doab. Vamsaka is the country of the Vatsa kings, of which Kausambi was the capital city. It apparently bordered on Madhyadesa. Virata² is the country in the vicinity of the modern Jaipur, wherefrom the Panchala country begins. The present town of Bairat is one hundred and fifty miles south of Delhi. Panchala is the Punjab, 'with a little territory in the more immediate neighbourhood of Hastinapura,' extending north and west from Delhi, from the foot of the Himalayas to the Chambal, Ahi-chhatra being the capital city of North Panchala or Rohilkhand, and Kampilya of South Panchala or the Gangetic Doab.3 The reading of the name of the country mentioned next is uncertain. It may be read as Gurjaka for Gurjara, as mentioned in the Panchatantra and the Rajatarangini, instead of Sphurjaka,4 and identified with the country of Gujarat. Kerala is the country of Malabar proper, on the western coast, extending farther down from Gujarat. Then comes Dravida or himavadvindhyayamamdhyam yatpragvinasanadapi | pratyageva prayagacca madhyadesah sa kirtinah || - ( Manu, II, 21 ). "It was at the court of the king of Virata that the Pandava princes and Draupadi passed the thirteenth year of their exile incognito. The Virata princess Uttara was married to Arjuna'ss on Abhimanyu, who, at the age of sixteen only, gallantly challenged simultaneously seven most famous generals of the Kaurava army at the battle of Kurukshetra. According to the Mahabharata, king Virata's capital was called Matsya, which Cunninghain finds in the neighbourhood of Jaipur. Wilson says: Dinajpoor Rungpoor, and Cooch Behar.' Apparently there was more than one country of this name and one would appear in Northern India. Manu (II, 19) places Matsya in Brahmarshi-desa. 3 According to the Mahabharata (Smith's History, p. 348), it would seem to have occupied the Lower Doab. Manu (II, 19) places it near Kanauj. Wilson will have it extending north and west from Delhi, from the foot of the Himalaya to the Chambal, and separated by the Ganges into Northern and Southern Panchala. Cunningham considers North Panchala to be Rohilkhand, with the capital city Ahi-chhatra represented by the ruins near Ramnagar; and the South Panchala to be the Gangetic Doab, with the capital city Kampilya between Budaun and Farrukhabad. 4 This term seems to mean literally something belonging to the first union of lovers, characterized by joy in the beginning and some expectation of fear in the end. Of the ten names, it should be noted, this is the only one which, as the name of a country, can be doubted, if the amended reading be not acceptable. 5 As applied to the classification of Brahmans (Pancha-Dravida, namely, Dravida Karnata, Gurjara, Maharashtra, and Tailanga), it has a much wider application embracing Gujarat, Maharashtra, and all the southern countries. 255

the country, where the Tamil language is spoken, extending from Madras to Cape Comorin. This tract is roughly bounded by the Vindhya range on the north, where Madhyadesa ends; Kerala or Malabar coast on the west; and Kalinga or Coromandal coast on the east. Kalinga implies the twelve-storeyed buildings of Kalinga,¹ the country along the Coromandal coast, north of Madras, wherefrom the Dravida country begins. It is clear, therefore, that India, comprising these ten provinces, extends from the Himalayas on the north to the Cape Comorin on the south, from Bihar, including perhaps North Bengal, on the east, to the Punjab and Gujarat on the west. A number of questions may now arise. Did these ten provinces cxist in a prosperous condition at any one time in the history of India? Did they ever possess gorgeous edifices of twelve storeys admitting of ten different types? How could the author of the Manasara come to know of them. Was the description of these buildings based on the details of the existing edifices, or was it meant to be an injunction to be followed in crecting edifices in these countries? Is there any reason to think that the Manasara is a technical treatise on architecture and not a work on polytechnics, like the Brihat-samhita, dealing casually with architecture and sculpture, nor an encyclopaedic work like the Puranas of northern India and the Agamas of southern India, which, too, have incorporated within them architectural and sculptural, as well as literary, religious, and scientific subjects ? That the Manasara is an avowedly architectural treatise meant. for professional students of architecture and written by an architect there need be no doubt. This will be clear beyond doubt even to a casual reader of this volume, not to speak of those who care to look up the writer's Encyclopaedia and Text or Translation of the Manasara. To me it is, further, clear that the Manasara was largely based on details gathered together froin the existing buildings and partly on details from the existing literature on the subject. It was, of course, meant to be a guide book, but it never aimed at being the sort of poem which is read for the beauty of its language or the general interest of its theme. It is very likely that the author of the Manasara was aware of the condition of buildings existing in the then India 1 The Calingae proximi mari of Pliny. jagannathatsamarabhya krsnatirantagah priye | alender: agarigayo: 11-(Tantras, see Apte's Dictionary). samprokto vamamargaparayanah utkaladdarsitapathah kalinagabhimukho yayau | - ( Ramayana, IV, 38 ) . - 256

comprising the ten provinces mentioned above. There might not have existed simultaneously buildings of twelve storeys in all the ten provinces. What seems to be really meant is the distinctive types of magnificent buildings belonging to each of these provinces.¹ For the Manasara is not a history of buildings of any country, it is a guide book, and as such it must give illustrations and generalize its findings. It matters not, therefore, if these countries were not equally prosperous at any time. It is sufficient that these countries had flourished and that they were well known in the history of India, and also that everyone of these could at some time or other claim prosperity and magnificence. Of these, Panchala and Dravidda are stated to be of the smallest types, next higher in size and importance are Madhyadesa, Kalinga, Virata, Kerala, and Vamsaka, the largest and most important are Magadha and Janaka. Sphurjaka is not specified (Manasara, XXX, 10-36). The last question to answer is how the author of the Manasara came to know of these provinces or divisions of India. If these provinces were autonomous and independent of each other the knowledge of them must have been received through literature, should a tour all over India for a purpose like this be thought an improbability in days before the establishment of the British Government. If, on the other hand, the internal affairs of all these provinces were settled by a common and central authority, who alone could decide upon a policy for common good, and under whom alone guide books like the Manasara discussing general methods and principles of buildings for all provinces could flourish, there must have been an empire comprising all or most of these countries. The probability of the latter view is strengthened by the consideration of the styles of architecture, apart from the types of buildings discussed above. These styles are also designated by geographical names, which imply much broader divisions, namely, Northern, Southern, and Eastern. They are called Nagara, Dravida, and Vesara. In the case of some architectural objects Vesara admits of two other branches, namely, Andhra and Kalinga, the three together constituting Tri-kalinga or three Kalingas. 1 About the existence of these types there need not be much doubt, because, for instance, Magadha, Panchala, Dravida, and others are used to imply types other than of buildings. For instance, Magadha stands for a Prakrt language? a tribe of people born of a Vaisya mother and Kshatriya father; Panchala stands for one of the four styles of composition; and Dravida for a language, a class of Brahmans, etc. The point is sufficiently claborated later on. 257 q

The Nagara style is distinguished by its quadrangular shape, the Dravida by its octagonal or hexagonal shape, and the Vesara by its round shape.1 • So far as is yet known, we cannot point to any buildings . . of very early date, or before the sixth or seventh century, if indeed quite so early. This is the statement made by authorities like Fergusson, Burgess, Smith, and others. This may be referred to all parts and all styles of India. Cunningham has gathered together fragments of what he calls the Gupta style, of which, however, no single example in its entirety can be cited. 6 The Dravidian temples generally consist of a square base ornamented externally by thin tall pilasters, and containing the cell in which the image is kept. In front of this may be added a mantapam or hall, or even two such, but they are not characteristic of the style. Over the shrine rises the sikhara, of pyramidal form, but 1 vedasram nagaram proktam vasvasram dravidam bhavet | suvrttam vesaram proktamandham syatu padakam || - (Manasara, XLIII, 124-25). This is applied to cars and chariots. The rules referring to buildings proper are given in XVIII, 92-104; XXVI, 76; XXI, 72-73, etc.; referring to sculpture proper, see for instance, LII, 78-100; LIII, 46-47 53-54, etc. dvarabhedamidam praktam jatibhedam gatah sranu | nagaram dravidam caiva vesaram ca tridha matam | kanthadarabhya vrttam yattadvemaramiti smrtam || grivamarabhya vastasram vimanam dravidakhyakam | na a agzi marm k 11-(Suprabhedagama, XXXI, 37-39). These also refer to buildings. For rules referring to sculpture, see the Kamikagama, LXV, 6-7, 12-18, and the Brihat-sanihila, LVIII, 4 (Kern's edition). "An interesting record from Holal is the label cut out on the capita of a finely carved pillar in the Amritesvara temple. It is called in the inscription a Sukara pillar. Speaking of the sculptor who made it, the record says that he, Bamunaja, the pupil of Padoja of Soge, was a Visvakarman, i.e., the architect of the gods in this Kali age, the master of the sixty-four arts and sciences, the clever builder of the sixty-four varieties of mansions, and the architect who had invented [?discovered] the four types [styles] of buildings, viz., Nagara, Kalinga, Dravida, and Vara. (Progress report of the Assistant Archacological Superintendent for Epigraphy, Southern Circle, 1914-15, p. 90.) $ In another inscription (Ep. Carnat., Vol. VII, Part I, Sorab Taluq, Inscription no. 275, Roman Text, p. 92, Translation, p. 46, note 1) these styles are called Dravida, Bhumijia, and Nagara,' of which Bhumija, which literally means, grown up on the spot,' may refer to the Vesara style, with Kalinga and Andhra as its two branches. These and many other quotations will be found in the writer's Encyclopaedia, under Nagara. Imperial Gazetteer of India, Vol. II, p. 171, para. 2. 258

: b PLATE XXXVII CHANDAKESAVA TEMPLE OF Belur, of star-SHAPED PLAN Page 258

PLATE XXXVIII Page 259 HOYSALESVARA TEMPLE, HALFRID

always divided into storeys and crowned by a small dome, either circular or polygonal in shape. Another special feature of these temples is the gopurams or great gateways, placed in front of them at the entrances to the surrounding courts, and often on all the four sides. In general design they are like the vimanas or shrines, but about twice as wide as deep, and very frequently far more important than the temples themselves. Another feature is the cornices of double curve; in other Indian styles the cornices are mostly straight and sloping downwards. As the contemporary northern styles are characterized by the prevalence of vertical lines, the Dravidian is marked by the prevalence of horizontal mouldings and shadows, and the towers and gopurams are storeyed. Then the more important temples are surrounded by courts enclosing great corridors or prakaras, and pillared halls.' < The square rathas were evidently models of Buddhist viharas and became the designs from which the temples proper or vimanas of Southern India were for long copied, and further, the oblong rathas, like Arjuna's temple, appeared to have given the first form to the great gateways or gopurams."1 Pierced stone windows are found at Ellora and other places. Regarding the Chalukyan style, which covers the Hyderabad territory, the Central Provinces, Berar, the Marathi-speaking and a part of the Kanarese-speaking districts of the Bombay presidency, it is stated that the earliest temples within this area, however, are not very clearly marked off from the Dravidian and the more northern style some of them have distinctly northern spires, and others are closely allied to the southern style.' For instance, 'the old temples of Papanath at Pattadakal presents a curious combination of styles. The body of the temple is Dravidian, but the Sikhara is a curious approximation to the form of the carly northern Hindu or Indo-Aryan order, while in details the temple shows a strong leaning to the Dravidian.' 'Still in Mysore, Dharwar, and Belgaum, as well as in Berar and Maharatha districts, sufficient remains still exist to illustrate the various development of the (Chalukyan) style." In the Chalukyan temples the corners are often made prominent by increments placed over them, or the whole plan is star-shaped, 1 Burgess cites (Imp. Gazetteer, II, pp. 172, 171) as examples of temples at Madura, Rameswaram, Tinnevelly, Srirangam, Kanchipuram, Pattadakal (Virupakhsha temple), Ellora (rock cut Kailasa temple). Burgess, ibid, p. 175. 259

the projecting angles having equal adjacent faces lying in a circle as in the temple of Belur in Mysore (built about A. D. 1120).' There are other examples where the Sikhara did not preserve the southern storeyed form but was rather stepped, forming square pyramid with breaks corresponding to the angles in the wall, and with a broad band answering to the larger face in the middle of each exposed side of the shrine.' 'The pillars are markedly different from the carlier Dravidian forms, they are massive, richly carved, often circular and highly polished. Their capitals are usually spread out, while middle section of the shaft is richly carved with mouldings in the round. They are almost always in pairs of the same design.' The richly carved and richly ornamented pierced windows belong specially to this (Chalukyan) style as we see it at Ajanta and elsewhere, just like the pierced stone windows employed in Dravidian temples at Ellora and other places. The buildings are erected without mortar, and the joints were carefully fitted. The whole was covered with sculpture, often of geometric and floral patterns, intermixed with numerous mythological figures, and, in the later examples, the courses of the base were carved with the succession of animal patterns prescribed for them in the Silpa-sastras. This is very fully exemplified in the great temple of Hoysalesvara at Halebid."1 These peculiarities of the Dravidian and the Chalukyan styles are taken from existing examples. Most of these details are also found under the Dravida style of the Manasara which, however, does not refer to the Chalukyan style as a separate order. The Northern or Indo-Aryan style of architecture covers the whole arca once occupied by the Aryans, usually designated as Hindustan,' to the north of the Tapti and Mahanadi rivers. 'What is known as the Jain style of architecture in Western India is a development or variety of this Indo-Aryan order, and was used by the Hindus and Jains alike all over Rajputana, Malwa and Gujarat. It was employed in its most ornate form by the Jains in their famous marble temples on Mount Abu, and by both the Jains and Hindus at Nagda near Udaipur. At Girnar also and Satrunjaya in Gujarat, as well as Khajuraho in Bundelkhand, are clusters of temples of this order."2 • Under this style are classified monuments of very various orders which may be separated into two or more distinct types.' The Vesara of the Manasara is apparently one of these orders. 1 Burgess, ibid, pp. 176, 177. ² Ibid, pp. 177, 179. 260

'The shrines and mandapas are square, and only slightly modified by additions to the walls of parallel projections, which, in the earlier examples, were thin, the walls were raised on a moulded plinth (pitha) of some height, over which was a deep base (adhishthana), the two together rising, roughly, to about half the height of the walls, over this is the paralleled face of the wall, usually of less proportionate height than in the Chalukyan style, and though devoted to figure sculptures in compartments, the tall thin pilasters of the southern style have disappeared, over this is the many-membered architrave, and cornice, above which rise the spire and roof. The spires follow the vertical lines of the wall, and present no trace of division into storeys, but vary in details with the age. In the earlier examples the summit was crowned by a large, fluted, circular block called amala (pure, shining) sila, probably mistaken for amalaka (Phyllanthus Emblica). The finial over this is of the shape of a vase, known as the Kalasa or Karaka.' 'One of the most striking features of the style is the richly carved domes over their mandapas or porches. (Nothing can exceed the elaboration and delicacy of details in the sculptured vaults of the temples at Abu and Nagda.) These, with the diversified arrangement of variously placed and highly ornamented pillars supporting them, produce a most pleasing impression of symmetry and beauty." The carlier examples were apparently astylar, then-like the southern forms with columns arranged in the mandapas in groups of four, and later, especially in Western India, the larger domes on twelve pillars formed the central area of the halls. These mandapas in early examples were roofed with long, sloping slabs, but, to provide for carved conical roofs inside, their outer forms represented courses of masonry, which were carved as in temples of Kanarak, Bhuvanesvara (older), Ambaranath, Baroli, Khajuraho, Abu and Chitor (mediaeval), Nasik, Benares, Udaipur, Satrunjaya, etc. (recent)." The peculiarities of the Nagara style, except in one or two rather unessential points, would correspond to these details of Northern or Indo-Aryan style.' The amala or amalaka-sila is not mentioned in the Manasara under this appellation, but the murdhni-ishtaka (brick at the top) seems to serve the same purpose as the amala- sila. The kalasa, or dome, sikha, and sikhara, are the distinguishing 2 Ibid, pp. 180, 181. 1 Ibid, pp. 178, 179. 261

features of the style found also in the Manasara in addition to the square shape. • 'The temples at Bhuvanesvara differ very markedly from those in the west in being almost entirely astylar-pillars having been introduced in later additions. They have the early form of sikhara-nearly perpendicular below, but curving near the summit, and the crowning member has no resemblance to anything like the small domes on Chalukyan spires.' Burgess, following the classification of Fergusson, has included the style found at Puri, Bhuvanesvara, and Kanarak under the Indo-Aryan or Northern style. But he has admitted that it may be separated into a distinct order.' What is called the Vesara in the Manasara seems to be identical with this style. The main characteristic feature of this style is, according to the Manasara, its round shape, and this is clearly exhibited by temples and images in the Orissan countries. The identification of Nagara with Northern India needs, however, an explanation. It seems to have been never before used exclusively in that sense. Moreover, it is the name of an extensive division in Mysore, a part in Tanjore, and a number of ancient villages in the Deccan. But it is found used more frequently as the names of villages, towns, and rivers in Bengal, Bihar, the United Provinces of Agra and Oudh, Rajputana, the Punjab, and Gujarat.2 Nagara is also the name of a portion of the Skanda-Purana, of a sect of northern Brahmans, and of a script. The Skanda-Purana, which to some scholars seems to have been named after Skanda-Gupta (A. D. 455-480), the seventh emperor of the carly Gupta dynasty, contains a part called Nagara-khanda. In this part of the Skanda-Purana it is claimed that the Nagara Brahmans are superior to all other Brahmans. It is held that they came over from the north and settled down in Gujarat at a place known as Nagranandana-pura. From these Nagara Brahmans, it is said, came the use of the Nagari alphabet, which belongs exclusively to Northern India. Indeed, it is very famous as the name of a script, particularly of Northern India extending from Bihar on the east to the Punjab and Gujarat on the west, 1 See the writer's Encyclopaedia, under Nagara. 2 J. A. S. B., 1896, Vol. LXV, Part I, pp. 116-117. Basu's collection of references in this Journal and many other quotations will be found under Nagara in the writer's Encyclopaedia of Hindu Architecture. 262

and from the foot of the Himalayas on the north to the Vindhya range on the south. This is the very tract which seems to have been covered by the Nagara style about the time of the Manasara. The southern and eastern tracts represented by the Dravida and the Vesara styles can also be associated respectively with the Tamil and the Telugu, including the Orissan scripts. As based on scripts and languages, these divisions, Nagara, Dravida, and Vesuru, have existed apart from the architectural styles. The expression Nagara is certainly not coined in the Manasara. Nagara is a common name for the town, and Nagara is an adjective therefrom and implies something connected with a city. Madura of Southern India is apparently an identical name to Mathura of Northern India. In the same way, the Nagarakhanda of Mysore, the port Nagore of Tanjore, and the village Nagara of the Deccan can be accounted for. There are several things to prove conclusively that the Aryan influence and civilization were spread, from Aryavarta or Northern India, all over the Dakshinatya or Southern India. It is true that the borrowed names sometimes became more prominent than those of which they are but imitations. New York of America, for instance, is much more prominent than old York of old England. Similarly the name Nagara, though originated in and indicating Northern India, might have become more prominent in Southern India. All these divisions are indicated by terms which were already in use as class names. The architecture of the country is divided into three broad styles and ten types, corresponding to the geographical divisions and the political entities. And there seems to have been a bond of union between these entities, however autonomous and independent they may have been in their mutual relation. In the total absence or rather non-existence of a unifying authority, the growth of a record of generalization, a guide book for the whole country, would be highly exceptional if not improbable. In other words, the presence of a standard work on architecture like the Manasara seems to pre-suppose an empire comprising countries having their own styles, methods, and principles, which are recorded and illustrated under so many divisions. Whether or not such books of generalization could have been written in those ancient days of India even without the patronage, active or passive, direct or indirect, of an imperial authority, it will be a useless, at any rate an 263

age of the MANASARA HINDU ARCHITECTURE IN unnecessary, discussion for our purpose. It is sufficient for me that the existence of such an empire may be taken to be conducive to the growth of such a treatise as the Manasara. Existence of an empire at the time of compilation of the Manasara seems to be indicated also by the following facts. In connexion with construction and disposition, according to ranks, of royal palaces, thrones, and crowns, royalty is divided into nine classes. They are called, in descending order, Chakravartin, Maharaja or Adhiraja, Mahendra or Narendra, Parshnika, Pattadhara, Mandalesa, Pattabhaj, Praharaka, and Astragrahin. The number of storeys and halls in a palace, the divisions of the whole compound into different courts, the quarters for royal personages and officials, and other buildings which are necessary adjuncts of an Indian palace, are described. The royal qualities, courts, army, and rate of revenue in accordance with the class to which a king belongs, are also incidentally mentioned. A consideration of these matters might have helped us in arriving at a time in the history of ancient India, if the historical materials and especially chronological data were available. What, however, concerns us most here is to ascertain the relation existing between these nine classes of kings. They are mentioned by their common names, and not by proper and personal names. An empire, in any case, has been expressly recognized in the Manasara. It is clearly declared that the empire of the Chakravartin or universal monarch reaches as far as the four occans. So it must include the whole of India, divided into three divisions, Northern, Southern, and Eastern, otherwise apparently known as Nagara, Dravida and Vesara. According to another classification we have seen, this empire seems to have comprised ten kingdoms. But here the empire is stated to have nine kinds of rulers. The Chakravartin is the suzerain of all the subordinate kings who send up tributes and taxes to him.3 The next king, called both Maharaja and Adhiraja, is the lord of seven kingdoms. Mahendra or Narendra 1 Manasara, Chaps. XL, XLI, XI, see pp. 124-126, 104-106 of this volume. 2 catuh sagaraparyantam mahim svikrtya balavan | jitva dvarasya purato ghamtamavadhya samsthitah | - (M., XLII, 6-7). 1 3 sarvavanindravanyosau cakravartiti kirtitah | - ( M., XLII, 10). evam ksudrasca bhupalah sve sve janapade karan | in svikrtya cakravartyadirajanam ca kurvantyapi || ( M., XLII, 75-76) • M., XLII, 11-13. 264

is the master of three kingdoms and more honourable than the Parshnika, Pattadhara, Mandalesa, and Pattabhaj classes of kings.1 The Parshnika is responsible for the administration of one kingdom, and the Pattadhara governs only half a kingdom.2 The rest seem to be chiefs rather than kings, though they possess their own army and courts. The Mandalesa is stated to be content with a mandala or province, while half a mandala or province is left to the charge of the Pattabhaj. The Praharaka is the king of several janapada or divisions, and the Astragrahin looks after several districts and is the ruler in a large city.4 About the Adhiraja it is stated that he must belong to the solar or the lunar race. 5 The kings of these races are Kshatriya by caste. Nothing is specified regarding the caste or castes of the other classes of kings. But the Praharaka is expressly stated to be born in a Brahman, Kshatriya, Vaisya, or Sudra family. This state of things points to a time when the Sudras were also recognized as kings. There does not seem to be much doubt that the recognition of these divisions presupposes the existence of an empire, the extent and the boundaries of which are made clear by the geographical classification of the ten types of gorgeous buildings and the three styles of the architectural and sculptural objects. These various divisions seem to represent the different schools of one system, the different branches of one united civilization and culture. For such an empire it is not absolutely necessary to find out a political head who can keep together the apparently separate and exclusive entities under his direct military control. When was there such an empire existing in India embracing the Nagara, the Dravida, and the Vasara portions all within itself? It is true, perhaps, that even in the time of Manu tracts of the country south of Vindhyas were known to the Aryans, and truer still that in the time of king Asoka, who partially conquered a portion of what we are now describing as Vesara and Dravida, there was a friendly intercourse subsisting between the north and the south. But the south was south still, and did not come to be considered as forming 1 M., XLII, 14-15. 2 M., XLII, 18, 21-22. 3 M., XLII, 23-28. '> * M., XLII, 29-35. 5 M., XLII, 12-13. 6 brahmanaksatriyavaisyasudranamekakulodbhavah | - (M., XLII, 29). 265

along with the north part of one and the same whole. The idea of such an empire as would include the whole of India from the Himalayas to Cape Comorin, from Gujarat to Bengal, had not yet grown up. It was still to come, and arrived much later when all the different parts came to be united under one hand. This leads us to consider next the course of development of such an empire, the story of which as a matter of fact is the story of the gradual spread of the Aryan influence and power from Aryavarta or Northern India southwards. Dr. Buhler seems inclined to think that the Aryan conquest of South India took place a considerable time before the Vedic period came to an end, and it certainly was an accomplished fact long before the authentic history of India begins at the end of the fourth century B.C.¹ According to Rhys Davids, till about the time of Buddha, Kalinga and part of the Deccan below the banks of the Godavari were outside the area of Hindu settlement.2 King Asoka conquered Kalinga and annexed it to his empire. The same monarch in his edict (XIII) refers to the Cholas, Pandyas, and Keralaputras as his pachantas or neighbours. It is evident from this edict of king Asoka that the three South Indian powers-the Cholas, Pandyas and Keralaputras-were, till the third century B.C., quite independent of Magadha. But presumably the friendly relation which had existed between king Asoka and those three powers opened for the first time the road of an exchange or amalgamation of two distinct civilizations, namely, the Aryan and the Dravidian. It is also not inconceivable that in or before the third century B.C., Andhra or Telugu country was in part Aryanized. And, lastly, it is clear from the Allahabad pillar inscription of Samudragupta that this Indian Napoleon directed his campaigns against eleven kings of the south, nine named kings of Aryavarta, besides many others not specified, the chiefs of the wild forest tribes, and the rulers of the frontier kingdoms and republics. He had also diplomatic relation with very remote foreign powers. Although it is at present impossible to identify every one of the countries, kings, and peoples enumerated in the inscription, enough is known to enable the 1 Apastamba, S. B. E., II, pp. xxxvi and xxxvii. 2 Sulla-Nipata, 1011, see also Anguttara Nikaya, 1, 213, IV, 252, 256, 260, Vinaya Texts, II, 146. The account of Rama's advance up to Ceylon as given in the Ramayana reflects a travel rather than a conquest. 266

historian to form a clear idea of the extent of the dominions and the range of the alliances of the most brilliant of the Gupta emperors.' He conquered South Kosala in the valley of the Mahanadi, subdued all the chiefs of the forest countries, which constitute the tributary states of Orissa, and the more backward parts of the Central Provinces, Pishtapura, the ancient capital of Kalinga, the hill-forts of Mahendragiri and Kottur in Ganjam, the kingdom of Mantaraja on the banks of the Kolleru lake, Vengi between the Krishna and the Godavari, Kanchi to the south of Madras, Palakka in the Nellore district, Devarashtra or the modern Mahratta country, and Erandapalla or Khandesh. This would imply the whole of the Dravida country bounded by the Coromandal and the Malabar coasts. The only place left by Samudragupta for his son Chandragupta to conquer and to annex to the empire was Kathiawar in Gujarat. The dominions under the direct government of Samudragupta thus comprised all the countries of Northern India. It extended from the Hooghly on the cast, to the Jamuna and Chambal on the west, and from the foot of the Himalayas on the north to the Narmada on the south. Beyond these wide limits, the frontier kingdoms of Assam and the Gangetic delta, as well as those on the southern slope of the Himalayas, and the free tribes of Rajputana and Malwa, were attached to the empire by bonds of subordinate alliance, while almost all the kingdoms of the south had been overrun by the emperor's armies and compelled to acknowledge his irresistible might. The empire thus defined was by far the greatest that had been seen in India since the days of Asoka. He maintained diplomatic relations with the Kushan kings of Gandhara and Kabul, and the greater sovereign of the same race, who ruled on the banks of the Oxus, as well as with Ceylon and other distant islands."¹ We now see that it was not until the time of the Imperial Gupta dynasty that the kind of empire implied in such a work as the Manasara came into existence. It is not our intention to say, indeed, that before or after this there had been nothing in the shape of an empire. It cannot certainly be gainsaid that there was a flourishing empire under king Asoka. It cannot be denied either that there was an empire flourishing in the south independent of the Aryans, that of the Andhras, so far back as about the beginning of the Christian era. The Chalukyas also built up an empire after the fall of the early 1 V. A. Smith, History of India (1908), pp. 271-72. 267

Guptas and remained powerful till about the middle of the eighth century, when the government of the country passed into the hands of the Rashtrakutas for more than two hundred years. Harshavar dhana of Kanauj also built an empire which, however, did not include within itself the Dravidian countries. What appears clear to us is that not till the reign of Samudragupta (326-375) or until Chandragupta II (375-413) was there any one empire which comprised the whole land, including the Telugu- and the Tamil-speaking places. It is further clear from the Allahabad inscription of Samudragupta that some of his subordinated kings belonged to the Sudra caste. The next internal evidence to be considered is one regarding religion. This is illustrated in the Manasara by the indifferent treatment accorded to the Buddhists and the Jains, and also by the unusually dignified manner of addressing the Brahmans as the gods on earth (bhu-sura), and lastly by the predilection for Vaishnavism. Two separate chapters are, however, devoted to the description of the Jain and the Buddhist images.¹ The description of the Jain deities, ostensibly the main object of a chapter, is submerged in a lengthy discussion of the various measurements used both in architecture and sculpture. The twenty-four Tirthankaras or Jain apostles are referred to, but not specified. The whole description of the Jain images is disposed of in a few lines at the fag end of the chapter. The Buddhist images are also described in a very small chapter of eighteen lines only. The account of these images too is very meagre. Evidently the author had in mind solely the effigies of Buddha, not of other Buddhist deities. This slight seems to have struck the author himself. So he adds in conclusion that the rest should be in accordance with the directions given in treatises specially dealing with these images.² The Buddhists and the Jains have been mentioned, it is true, in connexion with all matters referring to people of different sects. But the indifferent treatment accorded to the followers of Buddhism and Jainism is clear beyond doubt. For instance, in connexion with the village scheme described in a chapter of five hundred and forty lines, only two lines are devoted to them. The slight is all the more prominent from the fact that rather unwelcome quarters are 1 Chapters LV, LVI, see pp. 144-147 of this Volume. 2 ³a¶ myw¶guiaaanhaaga: 1-(M., LVI, 18, the last line). 268

reserved for the Buddhists and the Jains, and that the temples of their deities are built outside villages and towns.1 Similarly in connexion with buildings of different storeys they are treated with indifference, and nothing is specified about them.2 The same treatment is also apparent in connexion with the temples of attendant deities. The Buddhist and the Jain temples are passed over with the remark that they should be built according to the rules. of their own Sastras. It is true, however, that Buddha is recognized as one of the ten incarnations of Vishnu, whose family consists of the three groups of eight, sixteen including Buddha, and thirty-two deities. Again, in the chapter on pavilions (mandapa), which consists of five hundred and seventy-six lines, only one is given to the Buddhists and the Jains. s In connexion with the description of cars and chariots, it is stated in only one line that there should be one to seven platforms in the cars of the Buddhist and the Jain deities.4 Thrones and seats for the Buddhists and the Jains are left undescribed with the remark that they are ' thus stated.'6 1 bauddha pade caiva nairrtye tu jinalayam | - (IX, 387). durga ganapati caitra bauddha jainam gatavyam | anyesam panvadinam sthapayennagaradbahih || - (IX, 405-6) . evam tu coktavatkrtva bauddhadijitakalayam | tatamanastu tattadevannyametgat || -- - (XIX, 252-3, two lines out of 263 lines). devata disi castato vyasedvisnorovarajinaderalaye | - (XXI, 73-74, last two lines ) . ksudramanyamaya mukhyam bhaunake rudra visnujinakadiharmyake | -- (XXII, 98-99 ) . 3 See the summary, Chap. XXXIV, pp. 119-129. Compare the following: nairrtam tu bauddha svajjayante pautrikalayam | bauddham ca jinake caiva tatacchastrokta margavat | bauddhasyalaye tajjinalayam 'pi ksudralaye sarvasah | sastrasyalayatarana kathitam tatparivarakam kurvita || - ( XXXII, 149, 157, 165-6). 4 visnostatryambakasyapi caikadye (di ) navavedikah | bauddhadijinakantanam saptanam saptantam cakaditah || -- (XLIII, 144-5 ) - ' visnurudra jina kendra mukhyakanam sarvadevaganacakravatinam | asanani kathitani tani vai cotsavani kathitani suribhih || -(XLV, 211-112, last two lines). 269

Lastly, in connexion with the general description of images, the Buddhists and the Jains are left unspecified with a similar remark as before.¹ These are all the instances where the Buddhists and the Jains are at all mentioned. A significant point of omission also may be considered. Monasteries and such other architectural objects as are intimately associated with Buddhism and Jainism have not been referred to, while the minute details of Brahmanical Hindu temples have been rather elaborately described. From all this, two points seem to me to be clear. First, the Buddhists and the Jains, at the time of the Manasara, were not in a flourishing condition, secondly, they were not persecuted either. It was apparently a time of toleration for them. The next point to be clear about, is, which religion had the preference? It was Vaishnavism. The following references will, I hope, confirm this view. In support of the indifferent treatment accorded to the Buddhists and the Jains, the passages quoted above contain references to Saivism and Vaishnavism also. Vishnu and Isvara, Vishnu and Rudra, Vishnu and Tryambaka, and Brahma, Vishnu, Rudra are mentioned alongside Buddha and Jina.2 From this it must not be concluded, however, that Brahma, Vishnu, and Siva are treated in the same way as Buddha and Jina. In these passages it is directed how the latter should be treated, the former having been elaborately described. But in the treatment of Brahma, Vishnu, and Siva themselves a clear distinction and predilection have been shown. It is true that the opening verse is an invocation to Brahma, not to Vishnu or Siva, and that in the next verse the ultimate sources of the Silpa-sastras, like many other Sastras, have been ascribed to Siva, Brahma, and Vishnu.3 These deities are described in the usual order of Bralima, Vishnu, and Siva in the chapter dealing with the images of the Hindu Triad.4 1 a unforgrani gata farsza a | Medi a aj anyesam na pratimamevam manam tu samgraham || - (XLXIV 91 - 2 ) 2 See p. 269, notes 2, 3, 4, 5. 3 Manasara, I, 1, 2. gagari, kamalabhu, kamaleksana, this is the order ; but in Sanskrit the order may be changed. Here, however, the terms form component parts of a dvanda compound where the order of terms has some significance. 4 Chapter LI, see pp. 137-139. 270

n describing the riding-animals (vahana) of the Triad, the same order has been followed, the goose, the garuda bird, and the bull being treated in turn.1 It is also true that the phallus of Siva and his pedestal (pitha) have been elaborately treated in two separate chapters.2 This, however, does not seem to have been due to the author's or his patron's predilection for Saivism. For the phallus of Siva is a very famous object of Hindu sculpture, and it would have been given the prominence all the same even if the artist had belonged to an entirely different sect, because without this his treatise would have been incomplete. Similarly the extollation of the phallus worship added in conclusion may be explained.³ Preference for Vaishnavism seems clear also from the following points: The whole compound of a large building is divided into five courts around which the temples of attendant deities are built. Brahma, Vishnu, and Siva may individually possess attendant deities. There are, therefore, no reasons why the attendant deities of any one of the Triad should be specially treated, unless the author were closely in touch with the temples of any one group of the attendant deities wherefrom his ideas and illustrations originated. In this connexion the groups of eight, sixteen, and thirty-two deities of the Vishnu family alone are illustrated. The ten incarnations of Vishnu are also dealt with. But no mention is made of the attendant deities of Brahma or Siva. 5 This omission is significant, all the more because the Manasara is avowedly a treatise on architecture. If the work had been compiled in a place where Saivism or Brahma worship was favoured, the temples of their attendant deities could not but have been described in this connexion. A similarly striking omission in connexion with the Siva temples is also noticed in another important matter, namely, the foundations. Foundations of buildings are divided into two classes-according as they belong to temples and to human dwellings. Of the residential buildings there are four classes of foundations according to the four 1 Chapters LX, LXI, LXII, see pp. 149-151. 2 Chapters LII, LIII, see pp. 140-142. 3 Chapter LII, see p. 141. 4 Chapters XXXI, XXXIII, see pp. 116-117, 117-119. C Chapter XXXII, see p. 117. 271

castes, Brahman, Kshattriya, Vaisya and Suudra. 1 Of temples, those of Vishnu and Brahma are illustrated. Siva is not mentioned at all in this connexion beyond the author's usual method of passing on with the remark that the others should be similarly done. Very little is authoritatively known about the places in India where Brahma worship was ever so much favoured as Vaishnavism in Northern India, and Saivism in Southern India. The author's predilection for Vaishnavism seems to be indicated by this point likewise. In the laying out of villages and towns also the Vishnu temples have been given preference. It is stated that the Vishnu temples may be built anywhere in the village under the innumerable epithets of Vishnu, such as Sridhara in the cast, Vamana in the south, Vasu deva, Adi-Vishnu or Janardana in the west, Kesava or Narayana in the north, Nrisimha, Gopala, Rama and others at the four corners. No such details are given regarding the Siva temples. It is simply stated that the Isa (Rudra) temples may similarly be built in the quarters known as Rudra-jaya, Apa-vatsya, Jayanta, Parjanya and such other quarters, which are by no means prominent places in the village.2 1 Chapter XII, see pp. 106-108. 2 Manasara, IX, 255 f., and 383, for instance : ayamdisu caturdiksu gramasyapi caturdini | (255) evam yathestadigdese visnudhisnam prakalat || athava bahirage tu cestadigvisnoralayam | indradisu caturdiksu visnusthanam tu raksase || purvake srivaram proktam daksinam vamanam tatha | pascime vasudeva va cadivisnum janardanam || uttare kesavam proktam narayanamathapi va | antah praguttaradesa visnumurtim yathestakam || pitrdevesako va yatha nrsimhalayam bhavet | agnikone yatha ramam gopalalayameva ca || mitre ca tritalam kuryatsthanakam cadibhumike | dvitiyam casanam proktam trtiyam sayanam bhavet || athava sthanakam cordhve sayanam mulakasthale | istadigvisnurmyanam dvaram kuryadvicaksanah || gramasyabhimukham visnum narasimham paramukham | But in the case of Siva it is simply stated: ise vatha jayante va parjanyasya pade'pi va | (273) evamisalayam kuryadgramasya tu paranamukham || 272

JAIN DEEPA STAMBHA o ་ ཆ་ དཔལ་སང་ ་ ང་ ་ས་ ་ ་ ་ PLAN JAIN PILLAR PLATE XXXIX Page 272

GUPTA DOOR PLATE XL(a) Page 273

n the case of towns, the Vishnu temples alone have been taken into consideration. In the capital cities, it is distinctly stated that the Vishnu temples should be built at the main entrance.1 Similar illustrations from the body of the Manasara can be multiplied. But the point seems to be clear. Vaishnavism appears to have been the leading religion of the place where the Manasara was compiled. The author himself may have had a personal preference for Saivism or even for Brahma-worship, but his patron or the influence under which the author was working apparently had a leaning towards Vaishnavism in all its various phases and aspects, including even Buddha as one of the ten incarnations of Vishnu. Buddhism and Jainism, though by no means favoured religions, were allowed to continue. The influence seems to be one of non-interference, a universal toleration, with special preference for Vaishnavism. In which period of the history of ancient India, then, could Buddhism and Jainism have got on alongside Brahmanical Hinduism? The state of things that is reflected in the generous treatment of the followers of different religions was possible only in the period from the fourth to the eighth or ninth centuries of the Christian era. For during the reign of Asoka in the third century B.C., and some time after, Buddhism was in a very flourishing condition, while after the eighth or the ninth century both Jainism and Buddhism were declining. During this period kings of three distinguished dynasties reigned in the country. The Gupta empire in its entirety dates from the 1 See also 383, etc. sarvesam nagaradinam bhedam laksanamucyate | nagaram rajadhaniyam kevalam nagaram tatha | puram ca nagari camva khetam kharvatameva ca | kubjakam pattanam caiva sibiram vahinimukham | sthaniyam dronakam caiva samviddham kolakam tatah | nigamam skandhavaram ca durgam castavidham bhavet || nagaradini samgramam proktadurgam ca sattamam | rastramadhye naditire bahupunyajanavrtam || madhye rajayutam caiva nagaram krtamisyate | tatragate nagaryantam yadi visnvalayam bhavet | rajadhaniti tannama vidvadbhirvaksyate sada | - (Manasara, X, 37-47). 273

reign of Samudragupta (326-375), more accurately from the reign of Chandragupta II (375-413). The seventh and last emperor of the Gupta dynasty is Skandagupta. The imperial authority of the Guptas perished with Skandagupta (455-480), and the empire broke up, although the dynasty continued till about the middle of the seventh century or perhaps a little later. The Chalukyas came into power in the south at the beginning of the sixth century, after the fall of the Guptas, and remained powerful till about the middle of the eighth century, when the government of the Chalukya dominions passed into the hands of the Rashtrakutas for more than two hundred years. After their fall, the Chalukyas again came into power.1 Harshavardhana (606-648) also built up an empire in Northern India about the time when the Chalukyas were powerful in Southern India. None of these empires, however, comprised the whole of India. Buddhism and Jainism could not have flourished alongside Brahmanical Hinduism under the Rashtrakutas. Some of the Rashtrakuta kings may have been in favour of Jainism, but none seems to have favoured Buddhism. 'Under them,' says Sir R. G. Bhandarkar, 'the worship of the Puranic gods rose to much greater importance than before. The days when kings and princes got temples and monasteries cut out of the solid rock for the use of the followers of Gautama Buddha had gone by, never to return."2 ' During the two centuries of the rule of the early Chalukya dynasty of Vatapi,' says Vincent Smith,³ great changes in the religious state of the country were in progress. Buddhism, although, still influential, and supported by a large section of the population was slowly declining, and suffering gradual supersession by its rivals, Jainism and Brahmanical Hinduism. The sacrificial form of the Hindu religion received special attention and was made the subject of a multitude of formal treatises. The Puranic forms of Hinduism also grew in popularity, and everywhere elaborate temples dedicated to Vishnu, Siva, or other members of the Puranic pantheon, were erected. The orthodox Hindus borrowed from their Buddhist and 1 The early Gupta kings, about 300-530, the later Guptas A. D. 535-720. The Chalukyas of Badami, A. D. 550-753. The Rashtrakutas of Manyakheta, A. D. 753-973. The Chalukyas of Kalyani, A. D. 973-1190. 2 The History of Dekkan, p. 208. 3 V. Smith, ibid, p. 386. See also Bhandarkar, ibid, p. 191. 274

ain rivals the practice of excavating cave-temples.¹ Jainism was specially popular in the southern Maratha country.' On the other hand, the history of the early Gupta dynasty has all the necessary features. The empire of the Guptas comprised all the countries and divisions indicated in the Manasara. Brahmanical Hinduism was the leading religion, but Buddhism and Jainism were tolerated. King Meghavarna of Ceylon was allowed to despatch a mission with valuable presents to king Samudragupta for permission to build a monastery near the sacred Bo-tree at Gaya. The reign of Chandragupta II, the son and successor of Samudragupta, is noted for the visit of the Chinese pilgrim Fa-hien, who, being a Buddhist pilgrim, necessarily saw everything through Buddhist spectacles. In his account, mention is made of a number of monasteries along his journey from the Indus to Mathura, in which neighbourhood he found twenty of these buildings. It is evident that, with a Brahmanical supreme government, Hinduism of the orthodox kind must have been far more prominent than his account would lead the reader to suppose." Fa-hien was never 'stripped by brigands, a misfortune which befell his successor Hiuen Tsang. Probably India has never been governed better. The government did not attempt to do too much but let the people alone, and was accordingly popular.' Though the sovereign was a Brahmanical Hindu, the tendency to the harassing kind of persecution, which a Buddhist or Jain government is apt to display, was kept in check, and liberty of conscience was assured.' During the long and rather obscure reign of the next emperor, Kumaragupta (413-455), also Brahmanical Hinduism was the popular religion. This is clear from the fact that Kumara, like his grandfather, celebrated the horse sacrifice, a ritual repugnant equally to Buddhism and Jainism. Both Skandagupta (455-480) and Narasimhagupta Baladitya (485-535) 'continued to pay their devotions to the 1 There is no reference in the Manasara either to cave-temples or rock-cut pillars. Nor have free pillars like those of Asoka been specially described in the Manasara. There is no reason to think that an author who gives particulars of all sorts of buildings found all over the country should have remained entirely ignorant of these wonderful architectural objects. Their omission seems to have been due to this the Manasara is not an history of architecture. It is a guide book and was intended to help professional architects. Architectural objects like the cave-temples, rock-cut pillars, and free pillars had no more use for architects, presumably because they had become out of date at the time of the Manasara. 2 V. Smith, ibid, pp. 292, 293. 275

Hindu gods, while exhibiting, like Harsha in the seventh century, a strong personal predilection for Buddhist doctrine."1 'Whatever may have been the causes, the fact is abundantly established that the restoration of the Brahmanical religion to popular favour, and the associated revival of the Sanskrit language, first became noticeable in the second century, were fostered by the western Satraps during the third, and made a success by the Gupta emperors in the fourth century. These princes, although perfectly tolerant of both Buddhism and Jainism, and in two cases personally interested in the former, were themselves beyond question orthodox Hindus, guided by Brahman advisers, and skilled in Sanskrit. An early stage in the reaction against Buddhist condemnation of sacrifice had been marked by Pushyamitra's celebration of the horse-sacrifice towards the close of the second century. In the fourth, Samudragupta revived the same ancient rite with added splendour, and, in the fifth, his grandson repeated the solemnity. Without going further into detail the matter may be summed up in the remark that coins, inscriptions, and monuments agree in furnishing abundant evidence of the recrudescence during the Gupta period of Brahmanical Hinduism at the expense of Buddhism, and of the favour shown by the ruling power to classical Sanskrit at the expense of the more popular literary dialects, which had enjoyed the patronage of the Andhra kings."2 It is, further, clear from coins, inscriptions, and monuments that Vaishnavism was the predominating religion during the Gupta period. And this is the state of religious affairs evinced in the Manasara, namely, a Brahmanical Hinduism with preference for Vaishnavism and tolerant of both Buddhism and Jainism. ' 1 Paramartha, Buddhist of the sixth century, who wrote the life of Vasubandhu, states that Vikramaditya of Ajodhya, who at first was a liberal patron of the Sankhya philosophy, which is considered to have a strong affinity to both Buddhist and Jain doctrines, was induced by the eloquence of the celebrated Vasubandhu of Peshawar to turn a favourable ear to the teachings of Buddhism and to patronise its professors with equal liberality. The Queen and Prince Baladitya, who afterwards, about A. D. 485, succeeded to the throne as Narasimhagupta, both became disciples of Vasubandhu, and Baladitya, after his accession, continued his favours to the Buddhist sage. The coinage and official inscriptions of the Gupta kings are so distinctly Brahmanical that these statements might cause surprise." But it is fully confirmed by Hiuen Tsang, who describes Baladitya as a zealous Buddhist.-V. Smith, ibid, p. 292; Takakusu, J. R. A. S., 1905, p. 44; Watters, I, 288. 2 Smith, ibid, p. 287. 276

The appearance of treatises like the Manasara during the period of the early Guptas seems to be indicated by other reasons also. Following the spread and consolidation of the Gupta empire under Samudragupta there came a time of peace and quiet, especially during the reign of Chandragupta II, favourable to the cultivation of art and literature, and an intercourse of ideas and thoughts between the different parts forming members of one empire. It was in this Gupta period that a general literary impulse was extended to every department. In this classical period of Indian history an all-sided improvement in arts, literature and science came to be achieved. It was, again, during this period that the Sutra style of literature began to give place to the classic style. It has been shown elsewhere that the language of the Silpa-sastras, represented by the Manasara, seems to be the meeting place of the two.¹ Sanskrit was gradually raised to the position, which it long retained, as the sole literary language of Northern India. "The literary revolution,' says Vincent Smith, 'necessarily was accompanied by corresponding changes in the art of architecture. The forms of buildings specially adapted for the purposes of Buddhist ritual dropped out of use, and remarkable developments in the design of the Hindu temple were elaborated, which ultimately culminated in the marvellously ornate styles of the mediaeval period, extending from the ninth to the end of the twelfth century.' 12 The external evidences, mainly based on a comparison between the Puranas and the Silpa-sastras, also point to the same conclusions. The reasons have been elaborated for the belief that there is a relation of indebtedness between the Manasara on the one hand, and, on the other, the Matsya-Purana, the Bhavishya-Purana, the Agni-Purana, and the Brhat-samhita. 'To the same age probably (Gupta period)' says Vincent Smith, 'should be assigned the principal Puranas in their present form." Bana, the author of the Harsha-charita, who wrote about A.D. 620, 'carries back the proof of the antiquity of the Agni-Bhagavata-, Markandeya- and Vayu-Puranas four centurics further back than Alberuni, who in 1030 gives the list of the eighteen Puranas as given in the Vishnu-Purana, having seen three of them himself.' 1 See Appendix, and pp. 211-214 of the writer's Indian Architecture (1 st Ed., 1927) 2 Smith, ibid, p. 288, also refers to the seven characteristics of the Gupta style of architecture (see pp. 195-6 of this volume). Cunningham, Arch. Reports, IX, 42, I, V, X, XI, XIV, XVI, XX and XXI. 277

ELEVATION RESTORED. NOTE -The Plan and Details for Column etc. have been taken from Cunningham's Report Vol XXI. SECTION kou ☐ SIDE ELEVATION COLUMN. SCALE 1 INCH -1 FOOT GUPTA TEMPLE. NACHNA-KUTHARA. 10 10 15 20 25 30 35 40 SCALE OF FEET PLAN. GUPTA TEMPLE Page 276 PLATE XL(b)

The discovery of the Bengal manuscript written in Gupta hand has assigned the Skanda-Purana to the middle of the seventh century on palaeographical grounds.¹ Many other early quotations from, or references to, the Puranas have been collected by Buhler, who points out that the account of the future kings in the Vayu-, Vishnu-, Brahmanda and Matsya-Puranas seems to stop with the imperial Guptas and their contemporaries.' 6 This last observation,' adds Vincent Smith, indicates that the date of the reduction of the four works named (including MatsyaPurana, which seems to be intimately connected with the Manasara) cannot be very far removed from A.D. 500, the imperial Gupta dynasty having ended about A.D. 480. Buhler speaks of 'future kings,' because all the historical statements of the Puranas are given in the form of prophecy, in order to maintain the appearance of great antiquity in the books, which in their oldest forms were undoubtedly very ancient."³ The Manasara seems, therefore, to point to the Gupta period in view of the accumulation of external and internal evidences, both political, religious, and social, namely, the date of the Puranas, the existence of an empire comprising the whole of India, the division of royalty into nine classes (including the Sudras also), the popularity of the Brahmanical religion with predilection for the Vishnu cult and non-interference and toleration of Buddhism and Jainism, a general impulse to arts and literature, the appearance of the peculiar Sanskrit of the Silpa-sastras, and characteristics of architecture and sculpture consisting mainly of the three styles and ten types of buildings. At the time of the composition of the Manasara the memory of the first seven Gupta princes seems to have been fresh in the minds of the people. An expression gupta-vimsa has been used in the 1 J. R. A. S., 1903, p. 193. 2 Ind. Ant., XXV, p. 323. Ibid, pp. 19, 20; compare also: • The Vayu-Purana in its present shape seems to be referred to the fourth century A. D., by the well-known passage describing the extent of the Gupta dominions, which is applicable only to the reign of Chandra Gupta I in A. D. 320-326.³ The Puranas seem also to have been known to the author of the Questions of Milinda (Milindapamha), who composed a part of the work where the first references occur, almost certainly earlier than A. D. 300. (S. B. E., Vol. XXXV, pp. 6, 247.) 278

anasara¹ to imply sapta-vimsa or seven and twenty. Gupta in the sense of seven seems to have been coined in the Manasara. Perhaps it was due to the great fame and some patronage to the Manasara of the early Gupta princes, consisting principally of seven kings. For after the death of Skandagupta in A.D. 480, the seventh king of the dynasty, the empire broke up the next princes, Puragupta, Prakasaditya, and Narasimhagupta Baladitya being but chiefs. These conclusions are, however, in an apparent conflict with certain other matters. Cunningham has gathered together fragments of the Gupta buildings, wherefrom he draws the following peculiaritics of what he calls the Gupta style. 'The chief characteristic features of the Gupta temples are : (1) Flat roofs, without spires of any kind, as in the cave temples. (2) Prolongation of the head of the door-way beyond the jambs, as in the Egyptian temples. (3) Statues of the rivers, the Ganges and the Jumna, guarding the entrance door. (4) Pillars, with massive square capitals ornamented with two lions, back to back, with a tree between them. (5) Bosses on the capitals and friezes of a very peculiar form like the Buddhist Stupas or beehives, with projecting horus. (6) Continuation of the architrave of the portico as a moulding all round the building. (7) Deviation in plan from the cardinal points.' None of these characteristics seems to be applicable in its entirety to the buildings described in the Manasara. Spires, or sikhara and sikha, as well as the kalasa, or domes, are the chief characteristic features of the buildings described in it. These seven characteristics would point to the antiquated period of structural architecture. And Cunningham himself admits the fact. tadrksa'stakavrddhih syad guptavimsaksayam bhavat | varam grahena vrddhih syatsaptabhiscaksayo bhavet | - ( Manasara, LXI, 32-33). This refers to the following shadvarga, a set of six formulas, with which any particular measurement must conform before it can be accepted: • remainder of length * 8 / 12 534724 vyaya yoni vara fafa or amsa • " breadth * 9 / 10 " length x 8 27 " breadth x 3/8 • • " circumference, thickness, or height x 9 / 7 • " C * 9 / 30 " C * 4 / 2 More details will be found in the writer's Encyclopaedia, under Shadvarga. 279

* The style is similar to that of the cave temples of Udayagiri, and of the structural temples at Eran.' 'The use of flat roofs would seem to show that these buildings must belong to the very earliest period of structural architecture. When the architect, whose work has hitherto been confined to the erection of porticoes in front of caves, was first called upon to build the temple itself as well as the porticoes, he naturally copies this only prototype, and thus produced in a structural form the exact facsimile of a rock-hewn cave.'1 This seems to explain away the main objection. What is designated as the Gupta style points really to buildings of much carlier periods. By the time the Gupta dynasty was consolidated, the methods and principles of architecture seemed to have considerably improved, the architect invented the use of domes and other ornaments over the 'flat roofs' copied in the earlier periods from the rock-hewn caves. In the Gupta period proper, as truly held by later scholars like Vincent Smith, remarkable developments in the design of the Hindu temples were elaborated which ultimately culminated in the marvellously ornate styles of the mediaeval period, extending from the ninth to the end of the twelfth century.' So the characteristics of the real Gupta buildings, notably those which existed under the Guptas and are discussed in the Silpa-sastras, would be different from those given by Cunningham. The buildings described in the Manasara would conform, we have seen, to the characteristics of the Indo-Aryan and ChalukyaDravidian styles recorded from the existing examples by both Fergusson and Burgess. But none of these extant examples belonged to a period earlier than the sixth or seventh century A.D. These were, however, not the first of their class: buildings of this class must have cxisted long before the sixth or seventh century, because the extant examples themselves clearly show that they have passed through different stages in their development. The next objection may be one concerning the Gopuras, Prakaras, and such other objects which have been exhaustively described in the Manasara. These are undeniably the peculiarities of southern architecure. This objection may be easily disposed of. There seem to be sufficient reasons to hold that the account of architecture in the Manasara has reference to buildings of all parts of India, comprising the northern, southern, and eastern styles. The southern style might 1 Cunningham, Archaeological Survey Report, Vol. IX, p. 42. Some drawings are given in this volume as well as in I, V, X, XI, XIV, XVI, XX, etc. 280

be as elaborately described as the northern or eastern, even when the Manasara was compiled under the patronage of a northern emperor. The mixture of styles or the preferential treatment of one style over the other may similarly be accounted for. The 'sporadic appearance of temples of a style removed from their proper area may be accounted for in various ways; great temples were constantly being visited by pilgrims on their way from one shrine to another, and the repute of any new fane was soon spread over all India, and thus, when a prince undertook to build a new temple, an architect (sthapati) of acknowledged ability might occasionally be sent for from the most distant province, and engaged to design the work which, of course, would be in his own style." In the very same way the author of the Manasara might have been sent for from Southern India to compose the standard work on Indian architecture. The last point to be considered is the mention of Manasara in the Da sa-kumara-charita as a king of Malwa. This king Manasara is the hero of a fiction. There are no doubt historical facts concealed in a fictitious work. But it is not easy to sift facts from fictions. Those who are, however, inclined to connect the Manasara with this king of Malwa would assign the treatise to the seventh century, because the author of the fiction, Dandin, would be 'contemporary of Bharavi,' who is mentioned in an inscription of A.D. 634,2 and also of Harsha who reigned from A.D. 606 to 648. Historical facts, as stated above, cannot generally be extricated adequately from the complexities of the fiction. Moreover, although some vague conclusion has been inferred from the circumstantial evidence about the period in which Dandin, the undeniable author of the Dasa-kumara-charita, lived, no such vague idea even is available about the period or periods in which the semi-historical incidents described in the Dasa-kumara-charita might have taken place. Besides, king Manasara of the Dasa-kumara-charita, it may be incidentally pointed out, was not the hero or even one of the chief characters of the fiction. He is stated, as pointed out above, to be the king of Malwa and a contemporary of king Rajahamsa of Magadha, who was the father of Rajavahana, the chief of the ten princes (dasakumara). In the Dasa-Kumara-Charita itself king Manasara is stated to have been engaged in a war with king Rajahamsa, that is all. 1 Burgess, ibid, p. 178. 2 Macdonell, History of Sanskrit Literature, pp. 329, 332. 281

There is in the fiction practically no direct or indirect reference made as to the nature of interest which king Manasara might have been in the habit of taking in literary or artistic matters. It must, however, be admitted that there were no real occasions for such a reference.¹ In this connexion another incident, too, may be advantageously taken into consideration: neither in the three styles mentioned in the treatise Manasara under three geographical names (Nagara, Vesara, and Dravida), nor in the ten types of buildings bearing, again, geographical names and provincial divisions (Panchala, Dravida, Madhyakanta, Kalinga, Virata, Kerala, Vamsaka, Magadha, Janaka, and Sphurjaka), is included Malava, which was, presumably, the capital city and provincial kingdom of king Manasara of the Dasa-kumara-charita. In the circumstances it would be really doubly unwarranted to take any decision about the possibility or otherwise of king Manasara's patronage or instrumentality in the production of the standard treatise on architecture which, as its title would seemingly indicate, might have been named after him. King Manadeva of Nepal who ruled in the sixth century and was a great builder would be a more historical patron of the Manasara if it could be associated with him. In view of these facts we venture to expect that the reader may be inclined to consider more seriously the other evidences which are undoubtedly more authenticated and substantial, including those regarding the connexion of the Manasara with Matsya-Purana (A.D. 450) on the one hand and the Brhat-samhita (s. D. 550) on the other. On this assumption we shall perhaps be justified in placing the Manasara before the Brhat-samhita and somewhere close to the Matsya-Purana. In any event, we venture to hold that the evidences submitted above would warrant the extension of the period of the Manasara from A.D. 500 to 700, if not from the time of Vitruvius or carlier.3 1 In his two recently discovered works, called the Avanti-Sundari-Katha in prose and the Avanti-Sundari-Kathasara in verse, Dandin, the author of the Dasa-kamaracharita, is held to be well learned in architecture of royal and divine structures.' (Proceedings of the Second Oriental Conference, 1922, pp. 194, 196; see also p. 252, of this Volume.) 2 See pp. 286-287. 3 Until, however, the identity of the real author of the Manasara is established, and the missing link connecting the Manasara with Vitruvius is discovered and definitely ascertained, it would not be quite possible to be more precise about the date of the Manasara. 282 *