ACHARYA, P.K., Indian. Architecture According to the Mānasāra-śilpaśāstra, pp. iv, 268, index: A Dictionary of Hindu Architecture, pp. xx, 861, index. Both printed in Allāhābād, published by the OXFORD UNIVERSITY PRESS, and without date (1927 or 1928).

These two volumes, the latter especially, are monumental works, and will be indispensable to every student of Indian architecture and realia. Only those who work along these lines will realise the great labour involved in the preparation of such books, especially when they are almost the first of their kind; the serious study of the Indian śilpa-śāstras has been too long delayed, and a warm welcome may be extended to the Professors’ undertaking. The author, nevertheless, has neglected a good deal of work that has been done in this field; surprising omissions in the references, for example, are Rao, Tālamāna, Jouveau-Dubreuil, Archéologie du Sud de I’Inde, and texts such as the Viṣṇudharmottara and Śilparatna. Moreover the author is too little, if at all, acquainted with the actual buildings; otherwise, indeed, he could not have remarked that the buildings and sculptures of the time when the text of the Mānasāra was composed “have all been destroyed,” overlooking the fact that sculptures and building of this and earlier periods survive in thousands, and that a very great deal of exact information about the early architecture can be gathered from the Śuṅga, Kuṣāna, and Āndhra reliefs. I have myself in preparation a work based on this early material, which can and necessarily will be very fully illustrated. Jouveau-Dubreuil had the immense advantage of a thorough knowledge of the actual architecture, and of personal contact with living sthapatis able to explain the meaning of technical terms; without these qualifications Professor Acharya has attempted an almost impossible task, for here book-learning, however profound, is insufficient.

The following notes, however, are meant to be a further contribution to the subject and an acknowledgment of the value of what the Professor has already accomplished, rather than further criticism.

As of most general interest I would call attention to the items Ābhāsa, Candra-śālā, Hasti-nakha, Kuṭāgāra, Likh, Liṅga, Nārāca, Tulā. I should also like to emphasize the fact that a study of the early use of the words which later appear as established technical terms in the Śilpa-śāstras is of great value for the study of architectural history. There is still very much to be accomplished in this direction.

Ābhāsa: together with ardha-citra and citrābhāsa, are completely misunderstood. Neither of these is a material, but as explained by Śrīkumāra, Śilparatna, Ch.64, vv. 2-6 (see my translation in the Sir Ashutosh Mookerjee Memorial Volume), and by Rao, Elements of Hindu Iconography, I, p. 52, citing the Suprabhedāgama, a method. Both the Mānasāra and Suprabhedāgama as cited by the Professor himself are perfectly clear on the point; as the matter is important, I quote the latter:

Sarvāvayava-saṁpūrṇaṁ dṛśyaṁ tac citram ucyate
Ardhāvayava-saṁddṛśyam ardha-citraṁ caiva ca (sic).
Paṭe bhittau ca yo (al) lihyaṁ1 citrābhāsam ihocyate (sic).

The mistake about ābhāsa has led to the extraordinary view (Dict. p. 65, l. 3) that ālekhya is also a material. Citra, in fact is divided into citra, ardha-citra, and citrābhāsa, respectively sculpture in the round, reliefs, and painting. In Indian architecture, p. 70, in the same connection sarvāṅgadṛśyamāna, rendered “quite transparent,” really means “in which all the parts of the body are visible.” Of course, there are many cases where citra. by itself is used to mean painting, but some of these need critical examination; for example citrāṇi maṇḍalāni of Culllavagga, V, 9, 2 does not mean “painted circular linings,” as rendered in S.B.E., XX, but simply “carved bowl-rests.”

Ādhāra: add the meaning, “reservoir,” Arthaśāstra, III. 8 (Meyer).

Adhiṣṭhāna,plinth: Mukherji, Report on the Antiquities of the District of Lalitpur, 1899, describes and illustrates the various parts and mouldings. A few diagrams of this kind would have greatly enhanced the value of the Dictionary.

Ājira: a courtyard, see Geiger, Māhavaṁsa, Ch. XXXV, 3 and transl., p. 246.

Ālambana-bāha: the balustrade, vedikā, of a stair-way, sopānā, Cullavagga, V, 11. Cf. hasti-hasta. Ālambana, per se, is the plinth of a railing or balustrade.

Ālekhya: not in the Dictionary. See above under ābhāsa. The working drawing, on cloth, for the Lohapāsāda is thus designated in the Mahāvaṁsa, Ch. XXVII, 10. Ālekhya-sthāna is a space left in a manuscript for the subsequent insertion of an illustration.

Ālinda: balcony, gallery. Cullavagga, VI, 3, 5, glossed pamukha==pramukha: ib. VI, 14, 1, described as hatthi-nakhakaṁ, see hastinakha. In Mahāvaṁsa, XXV, 3, the rendering of ālinda as “terrace in front of a house door” (Geiger, Mahāvaṁsa, p. 246, note 2) seems very questionable.

Āmalaka: not in the Dictionary, though discussed in the other volume, p. 179, where kalaśa, “vase” (finial) is misrendered “dome.”

Not in the Mānasāra, and the suggested equivalent mūrdhni-iṣṭaka seems a little questionable. I doubt if an example as finial could be cited before the Gupta period, when it can be seen on the reduced edifices of the Sārnāth lintel (Sahni, Catalogue, pls. XV-XXVI); but these imply an already well-established tradition. The form is already employed architecturally in connection with pilasters represented at Amarāvatī. In Cullavagga, VI, 2, 4 a kind of chair is termed āmalaka-vaṅṭika-piṭḥaṁ, and this is glossed by Buddbaghosa as “having large āmalaka-formed feet attached to the back.” The translation “many feet” of S.B.E. XX, 165, cited by Acharya without comment, can hardly be justified, though Buddhaghosa’s  bahupāda suggests it at first sight. Amongst the countless representations of chairs and couches in Indian art of all periods I cannot think of a single example with more than four legs.

Aṅgana: applied to the enclosure surrounding a stūpa, i. the circumambulation-platform between the stūpa and its railing, Dhammapada Atthakathā, 290 (Bk. 21, Story 1, Burlingame, H.O.S., vol. 30, p 175).

Aṇldvāra: Arthaśāstra, II, 3,.and III, 8. Meyer renders “sidedoor,” Shamatsastry “front door.” In III, 8, the latter meaning would seem to be indicated, as only one door is mentioned, and the window above it is referred to. In the early reliefs we see no side doors to ordinary houses, while there is generally a window above the single (front) door.

Aratni: add references to Kauṭilīya Arthaśāstra, II, 20, with a table of measurements practically identical with that of the Mānasāra. In Arthaśāstra. II, 5, the rain gauge (s.v. kuṇḍa below) is to be an aratni in width, i. 2 spans (vitasti) or 24 aṅgulas.

Argala: Pali aggala, Simhalese agula, a bolt. See under dvāra below.

Arghya: not in the Dictionary. In Mahāvaṁsa, XXX, 92, Geiger’s  rendering of agghiya as “arches” is impossible. Agghiya-pantī may be rows of garlands or swags, a. common enough ornament, or more likely rows of vessels of some kind; Phaḷikagghiya must be a crystal dish or platter, as it has four corners in which are placed heaps (rāsiyo) of gold, gems, or pearls—but more likely we should understand phalakagghiya and translate as “wooden offering table” or “altar.” In any case “four corners” has no meaning in connection with any sort of known torano. Agghika of Mhv. XXXIV, 73 is more doubtful, perhaps here equivalent to altars or reredos (Siṁh. wahal-kaḍ). See also agghiya, agghika in P.T.S. Pali Dictionary.

Āryaka-stambha: not in the Dictionary: but see under āveṣanin, below, and Dictionary, p. 669.

Āsandī, a throne, seat: Atharva Veda, XV, 3 (see Whitney, in H.0.S., Vol. VIII), where the various parts are named; the description suggests the types still seen at Amarāvati.

A detailed nomenclature of seats will be found in Cullavagga, VI, 2. Cf. ib., VI, 14, also Brahmajāla Sutta, (Dialogues, I, p. 11, note 4).

Pace S.B.E. XVII, p. 27, it is by no means demonstrable from Jātaka I, 108, that āsandi means “cushion”; Cowell’s  “couch” is undoubtedly correct, and this is the sense everywhere else. To suppose a chair or couch placed in a cart presents no difficulty.

Aṭṭāla: watch.-towers or gate-towers, Milindapan̄ha, V, 4. Gopuraṭṭhā, Mahāvaṁsa, XXV, 30. Gopuraṭṭālaga, Uttarādhyayanasūtram, IX, 18, Charpentier, pp. 97, 314.

Avasaraka: osaraka (Pāli) (? that which sheds water) overhanging eaves [of a building without verandahs, anālinda), Cullavagga, VI, 3, 5: glossed as chadana-pamukhaṁ, “projecting from the roof.” Osārake, “under the eaves,” i. outside the house, Jātaka;, 111, 446. Cf. modern chajja.

Āveṣaṇin: not in the Dictionary; architect, foreman. Inscription on Sāñcï south toraṇa, “Gift of Ānanda, son of Vāsiṣṭhi, āvesaṇin (rendered “foreman of the artisans”) of Raja Śrī Śātakarṇi” (Marshall, Guide to Sanci, p. 48). Āyaka (āryaka)-stambhas dedicated by Siddhārtha son of Nāgacanda, both āvesaṇins (Burgess, Notes on the Amaravati Stupa, p. 56); āvesa is stated to mean a workshop, atelier.

Ayas: not in the Dictionary. This word is always used for iron (see loha, below). Mahāvaṁsa, XXV, 28, ayo-kammata-dvāra, “iron studded gate” (of a city); ib., 30, ayo-gulaṁ, “iron balls”; ib., XXIX, 8, ayo-jāla, an iron trellis used in the foundations of a stūpa. Reference might have been made to the iron pillars st Delhi and Dhar, and the use of iron in building at Koṇārak.

Bodhi-ghara, mahābodhi-ghara: temples of the Bodhi-tree, presumably like the many examples illustrated in the early reliefs. No doubt a pre-Buddhist form, preserved in connection with the cult of the Bodhi tree. See Mahāvaṁsa, XXXVI, 55, XXXVII, 31, etc.; in the former place provided with a sand court, vālikātala; ib., XXXV, 89 (aṅgana. Also called a maṇḍapa, ib., XVIII, 63.

Bodhi-maṇḍa(la): is treated as synonymous with vajrāsana, but is really the special area within which the vajrāsana is established; see Hsüan Tsang as cited by Watters, II, 114, 115.

Candra(-śāla), etc.: some useful material is contributed towards a solution of the problem of the proper designation of the so-called “caitya-window” (dormer or attic window, gable, etc.), one of the commonest and most distinctive motifs recognizable in Indian architecture from first to last. “Caitya-window” is unsatisfactory, as the form is by no means peculiar to, nor can it have been originally devised expressly for caitya-halls; the gable form is derived from that of an ordinary barrel-vaulted house end. Toraṇa is perhaps correct in so far as the window is actually an arch, vātāyana in so far as it is a window, but neither is sufficiently specific. The problem is a little complicated by the fact that we have to do both with arched windows actually admitting air to upper chambers, dormers, or attics, with real internal space, and also with similar forms used decoratively and placed in series on cornices or similarly used in. friezes; but the various architectural forms, complete figures, or heads (see also gandharva-mukha, and gṛha) which appear framed in the niche formed by the window-arch prove that the idea of an opening to internal space is always present. The best established word is Tamil kūḍu (Jouveau-Dubreuil, passim), but there seems to be no similar word in Sanskrit; kūḍu means nest, and it applies both to the window as an ornament, and to actual pavilions (karṇa-kūḍu, Jouveau-Dubreuil, Dravidian Architecture, fig. 4). The proper term in Sanskrit seems to be candra-śālā (see s.v. in the Dictionary), meaning either a gabled chamber on or above the kapota (for which candra is given as a synonym), or the goble window itself. In the last case candraśalā should really be an abbreviation of candra-śāla-vātāyana, and this seems to be the most explicit designation: “gable-window” is probably the best English phrase, German dachfenster.

A number of passages seem to show also that gavākṣa may be synonymous with candra-śālā-vātāyana. Thus in Raghuvaṁsa, VII, 11, the gavākṣas are crowded with the faces of beautiful young women looking out, and ib. XIX, 7, Agnivarman is visible to his subjects only to the extent of hia feet hanging down from the gavākṣa. The modern vernacular equivalent is of course jharokhā.

The many-cusped arch, known to modern Musalmctn masons as piyālidar mihrāb, and familiar in Rajput, Mughal, and modern Indian architecture, is a development of the “horse-shoe” arch (gable window) which has rightly been regarded as of Indian, pre-Muhammasdan invention (Rivoira, Moslem Architecture, p. 110f); every stage in the evolution cas be followed. Cusped arches are found already in Java by the eighth century (Borobuḍur); there is an excellent example at the Gal Vihārē, Paḷn̄āruva, Ceylon. It would take too much space to treat this interesting subject at length here, but it is worth while to note that Mukherji, Antiquities of the Lalitpur District, I, p. 9, gives the Indian terminology; the “parts of the so-called Saracenic (five-foiled) arch, are all Hindu.” These names are, for the spring of the arch, nāga (cf. nāga-bandha. in the sense ot chamfer-stop); for the foils or cups, kaṭora; and for the top, cūkkā (?=cūlikā, q.v. in Dictionary).

Caṅkrama: cloister, monk’s  walk, at first perhaps only paved, later roofed and railed (Cullavagga, V, 14, 2, 3). Caṅkamana-sālā, “hall in a cloister,” Cullavagga, V, 14, 2 and Mahāvagga, III, 5.

Cetiya-ghara: in Mahāvaṁsa, XXXI, 29, and 60, 61, cetiya-ghara is a structure built over a stūpa, thūpaṁ tassopari gharaṁ. Some have seen evidence of such a structure in the still standing tall pillars surrounding the The Thūpārāma Dāgaba at Anurādhapura, and this interpretation seems to be plausible, especially as the pillars are provided with tenons above. An actual example of a stūpa with a roof over it, supportad by four pillars,  can be seen at Gaḍalādeniya, near Kandy, Ceylon. The old caitya-halls are also, of course, cetiya-gharas, and of these there existed also many structural examples. “

“Thūpaghara... is simply a house over a tope” (Hocart, A.M., Ceylon. Journ. Science, G., Vol. I, p 145).

Chan̄avira: some description might have been given of this very common ornament, found from pre-Mauryan times to the present day. See Rao, Elements of Hindu Iconography,I, p. xxxi, and M.F A. Bulletin, No. 152, p. 90. The chan̄avira passes over both shoulders and both hips, crossing and fastening in the middle of the breast and middle of the back; it is worn by deities and men, male and female, and occurs also in Java.

Citra: art, ornament, sculpture, painting, see above under ābhāsa. Citra, citra-karma do not always mean painting. Some places where the word occurs and has been so translated need reëxamination; for example, Cullavagga, V, 9, 2, citrāṇi maṇḍalāni does not mean “painted circular linings,” but rather “carved bowl-rests.” Some referencee should be given to citra-sabhā, citra-śāla which are of very common occurrence in the sense “painted hall or chamber.” The citta-sabhā of Jacobi, Ausgewählte Erzählungen, p. 39, has a high tower (uttunga siharā). Description of a citta-sabhā cited from the Uttarādhyayana Sūtra, Meyer, Hindu Tales, p. 174. Cittāgāra, in Sutta Vibhaṅga, II, 298.

Cūlikā: as something at the top must be connected with cūḍā. But in Mānasāna, L. 301, (Dict., p. 197), lamba-hāram api cūlikādibhiḥ, cūlikā must be “bodice,” and synonymous with coḷaka.

Daraninavami-śilā: not in the Dictionary. A square stone (or rarely bronze) slab or box divided into nine compartments in which are placed symbols connected with water, the whole being laid below the foundations of a temple or below an image (A.S.I., A.R., 1903-04 p. 98, note). This object is known in Ceylon as a. yantra-gala, where several examples have been found (Parker, Ancient Ceylon, pp. 298, 658; Mem. Colombo Museum, Series A, I, p. 25).

Deva-kula: in the Avadāna-śataka (Feer, p. 98), used of a temple of Nārāyṇa. See also A.S.I., A.R., 1911-12, p. 124. Devakula of the Nāga Dadhikarṇa, Mathurā inscription, Lüders’ List, No. 63. Inscription of Loṇāśobhikā on Mathurā āyāgapata, see VI Int. Congr. Orientalists, III, p. 143.

Dhavala, whitening: applied to a plastered or other surface, śilparatna, Ch. 64. Dhavala-hara, a “White House, “ palace, Haribhadra, Sanatkumāracarita, 548, 599, 608.

Drupada: a post, Rg Veda, 3, 32, 33. The whole passage is very doubtful, but apparently two horses are compared to carved figures of some kind (brackets?) upon a wooden post.

Dvāra: the parts of a door are listed in Cullavagga,  V, 14, 3, also ib. VI, 2 (not quite correctly translated in S.B.E., XX, p. 106), as follows: kavāṭa, the leaves; piṭṭhasaṁghāṭa2 (=Sanskrit prasthā-saṁghaṭikā, “upstanding pair”), the door-posts; udukhallika, threshold;  uttarapāsaka, Iintel; aggalavaṭṭi,  bolt-post; kapi-sīsaka, bolt (-handle);  sūcika, the pin or part of the kapi-sīsa which fits into the socket in the bolt-post (cf. sūci=cross-bar of a vedikā);  ghaṭikā, apparently the slot in the bolt-post  just referred to; tālacchidda, key-hole;  āviñchanacchidda, string-hole; āviñchana-rajju, string for pulling the leaves to from outside preparatory to locking. Some of these terms occur elsewhere; with reference to a passage in the Mahāparinibbāna Sutta where Ānanda leans against the kapi-sīsaka Buddhaghosa is certainly right in glossing kapi-sisaka as aggala, for the Siṁhalese agula is big enough to lean against (see my Mediaeval Sinhalese Art, figs. 80-82, for illustrations, ib. p. 133, for the Sinhalese terminology). As in so many other cases the terms are perfectly comprehensible when the objects have been seen as represented in relief, or in use, and when the modern technical terms are known.

As correctly observed in S.B.E., XX, p. 160, dvāra is “doorway,” “aperture,” always with reference to outer doors or gates of any building, or of a city, while kavāṭa means the leaves of a door, the door itself.

See also under gṛha, and cf. Robert knox’s  description of the palace of Rāja Siṁha II, “stately Gates, two-leaved ... with their posts, excellently carved.”

Bahi-duāla-sāIā=bahir-dvāra-śālā, “outer room,” “gate chamber,” Mrcchakaṭika, III, 3.

From RV. I, 51, 14 we get duryo yūpaḥ for the door posts, from RV. I, 113, 14 ātā for the door leaves, and from RV. III. 61. 4 a thong (syūman) fastening.

Dvāra-bāhā: door posts, Mahāvaṁsa, XXV, 38: ayo-dvāra, ayo-kammatadvāra, ib. XXV, 28, 29, 32.

Dvāra-koṭṭhaka, gate house: cittakūṭa dvāra-koṭṭhaka, etc., “a gate-house with a decorated peak, and surrounded by statues of Indra, as though guarded by tigers,” Jātaka, VI, 125: cf. Dhammapada Atthathā, Bk. 2, story 7.

For koṭṭhaka see also Cullavagga, V, 14, 4 and VI, 3, 10; Jātaka, I, 351 and II, 431; and Meyer, Arthaśāstra, p. 75, note 5 (in the sense of “shrine”) . Koṭṭhaka is usually “gatehouse, “ but piṭṭhikoṭṭhaka is “back-room” in Dhammapada Atthakathā, II, 19.

In Jātaka I, 227, dvāra-koṭṭhaka is, as usual, gate-house, not as interpreted in S.B.E. XVII, 219, ‘mansion’ (the ‘mansion’ is ghara and it has seven dvāra-koṭṭhakas).

Gāirikā: red chalk. Cullavagga, V, 11, 6, geruka, red coloring for walls. Medium red color, Śilparatna, Ch. 64, 117. Brown, Indian painting under the Mughals, p. 124 (used in preprtring the lekhanī or pencil). Used as rouge, Karpūramañjari, III, 18, see H.O.S., Vol. 4, note on p. 268. as a pigment, dhātu-rāga, Meghadūta, 102. Geruka, Culllavagga, V, 11, 6, VI, 3, 1, and VI, 17, 1. Mahāvagga, VII, 11, 2.

Gaṇḍa-bheraṇḍa: insufficiently explained by the cross-reference to stambha. The two-headed eagle, a gigantic bird of prey, is first found in India on a Jaina stūpa base at Sirkap (Marshall, Guide to Taxila, p. 74) . In mediaeval art two forms appear, analogous to those of garuḍas, one with a human body and two bird heads, the other entirely bird. Connected especially with the kings of Vijayanagar, and appearing on their coins, carrying elephants in its claws. Other examples at Śriśāilam (A.S.I., A.R., Southern Circle, 1917-18); remarkable panels st Koramaṅgala and Belūr, Cāḷukyan (Mysore A.S.Rep., 1920, and Narasimachar, Keśava temple at Belūr, p. 8). A common motif in south Indian jewellery. In Ceyion, see my Mediaeval Sinhalese art, p. 85. Cf. also hatthilinga-sakuṇa, Dhammapada Atthakathā, 1, 164. Further references will appear in the Boston Catalogue of Mughal Paintings.

Gandha-kuṭi, see s.v. Kuṭi.

Gandharva-mukha: designation of the busts or faces framed in the openings of kūḍu, candra-śālā-vātāyana, or gavākṣa, gable windows (Jouveau-Dubreuil, Dravidian Architecture, p. 12). Cf. canda-muha, s.v. candra-śālā.

Gavākṣa: see Candra, Gandharva-mukha, Gṛha, and Harmya.

Gṛha, ghara, āgāra, geha, etc.: there is an excellent description of Vasantasenā’s  house (geha, bhavana) in the Mṛcchakaṭika, IV, 30, seq. There are eight courts (paoṭṭhā=prakoṣṭha);3 above the outer door (geha-dvara) is an ivory toraṇa, supported by toraṇa-dhraṇa-tham-bha, and stretching up its head (sīsa) towards the sky; at each side are festival jars (maṅgala-kalasa)--“Yes, Vasantasenā’s  house is a beautiful thing.” In the first court are pāsāda-panti, rows of pavilions, having stairways (sobāṇa), and crystal windows (phaṭi-vāḍa=sphaṭika-vātāyana) with moon-faces (muhe-cande), or probably “faces on the candra, “ i. gandharva-mukhas framed in the candraśālā-vatāyanas ornamenting the roll-cornice, for which the description “seeming to look down upon Ujjayinī “ would be very appropriate. In the third court are courtezans carrying pictures painted in many colors, vivihavaṇṇi-āvalitta citraphala=vividhavarṇikāvalipta citraphalaka. In the fourth court, where music and dancing take place, there are water-coolers (salila-gagario=salilogargarayaḥ) hanging from the ox-eye windows (gavekkha=gavākṣa).

Tisalā’s palace in the Kalpa Sūtra, 32, is a vāsa-ghara, dwelling place; it is sacitta-kamme, decorated with pictures, and ulloya-cittiya, has a canopy of painted cloth (cf. Pāli ulloka).

Milindapañha, II, 1, 13 has “As all the rafters of the roof of a house go up to the apex, slope towards it, are joined together at it.”

The famous triumph song of the Buddha (Nidānakathā, Jātaka, 1, 76=Dhammapada, 154) has “ Broken are all thy beams (phāsuka), the housetop (gaha-kūṭa) shattered”: the housebuilder is gahakāraka.

See also Bodhighara, Cetiyaghara, Cittāgāra, Dhavala, Kūṭāgāra, Samudrāgāra, Santhāgāra.

Harmya: ramyaṁ harmyam, a beautiful palace, Vikrama Carita (Edgerton, text and transl. in H.O.S. 26, p. 258, and 27, p. 239) has the following parts: mūlapratiṣṭhāna, basement; bhitti-stambha-dvāratoraṇa, walls, pillars, doorways and arches; śālabhaṁjikā, statues; prāṅgaṇa, courts; kapāṭa, folding doors; parigha, door-bars;4 valabhi, roofs; viṭaṅka, cornices; nāga-danta, pegs; mattavāraṇa, turrets; gavākṣa, ox-eye windows; sopāna, stairs; nandyāvartādi-gṛha, pavilions (? ) (see Dictionary, s.v.). Harmikā, the little equare structure on the top of a stūpa. (Divyuvadāna) . A cross reference to rāja-harmya should be given in the Dictionary.

Harmya, dwelling, Atharva Veda, XVIII, 4, 55; RV. I, 121, 1, I, 166, 4, VII, 56, 16, etc.

Savitāna-harmya, Raghuvaṁsa, XIX, 39, “place with an awning”; or perhaps vitāna=modern chajja.

Hasti-hasta, gaja-hasta: amongst innumerable examples might be cited one at Nārāyaṇpur, Burgess, A.S.W.I., III, pl. XXXI, 3. Elephant-trunk balustrades in Ceylon are ẹt-hoṇḍa-vẹl, with the same sense as hasti-hasta.

Hasti-nakha: literally “elephant’s nail.” In Cullavagga, VI, 14, 1 a pāsāda having an ālinda (balcony, gallery), qualified as hatthi-nakhakaṁ, is a permitted monastic residence. According to Buddhaghosa’s gloss this means hatthi-kumbha patiṭṭhitaṁ, literally “ supported on elephants’ frontal globes,” and so to be rendered “supported by pillars having elephant capitals”; and this is plausible enough, as pillars with elephant capitals, supporting galleries and upper storeys, are highly characteristic of early Indian architecture. It is true that one hesitates to accept nakha in any other sense than that of “nail” or “claw.” But it is possible to retain the interpretation “elephant capital” without supposing that nakha=kumbha, for in fact the observer, standing at the foot of such columns, e.g.  at Beḍsā (see accompanying Plate), and looking upwards, sees nothing of the actual capital, except the under sides and nails of the fore feet of the elephants, which project beyond the edge of the abacus, and this may well have given rise to the term “elephant’s nail” as applied to elephant capitals.

On the other hand, hasti-nakha occurring in the śiśupālavadha, III. 68, Śanairaniyanta, III. 08, Śanairaniyantra rayāpatanto rathāḥ. kṣitiṁ hastinakhāt... turaṅgaiḥ, “the swift chariots are slowly brought down from the hastinakha to earth by the horsee,” seems to refer to a place or structure on the rampart. Amara’s gloss is pūrdvāri mṛtkūṭaḥ “a kūṭa made of earth at the city gate.”

The word also occurs in Kauṭiliya Arthaśāstra, p. 53 of Shamasastry, the Dictionary citing only Shamasastryś translation s.v. gṛha-vinyāsa. Here too, hasti-nakhas are connected with the gate and rampart of a fort. Meyer’s version, p. 71, given here with slight modification, is much to be preferred: “For access, an ‘Elephant’s nail,’ level with the opening of the gateway, and a drawbridge (saṁkramaḥ saṁhāryo); or in case there is no water (for a moat), a causeway made of earth.” The hasti-nakha is here then presumably a pillar with an elephant capital, standing in the moat, to receive the drawbridge when the latter is let down upon it, or pushed out onto it.5 It is not impossible that the term hasti-nakha, by an extension of the original and strict meaning, had come to be applied also to the drawbridge itself, and even to the causeway.

The śiśupālavadha passage would then imply simply the bringing of the chariots across the drawbridge, or, as understood by Amara, across the causeway of earth which takes its place when there is no water; and thence onto the solid ground.

Cf. Keśanakha-stūpa, s.v. Stūpa, not explained (Feer, Avadāna śataka, p. 487), but possibly with some reference to a lion capital.

Hasti-prākāra, see Prākāra.

Hasti-prstha gaja-pṛṣṭha: this appropriate name is applied to the buildings with apsidal structures, common in Pallava, Cola, and later Dravidian work (see accompanying Plate). The  reference on p. 159 to Indian Antiquary XII should be corrected to XL. On p. 398 hastipṛṣṭha single-storeyed buildings are said to have an “oval steeple”; read instead “apsidal roof.” The Professor elsewhere often refers to oval buildings, perhaps meaning apsidal; an oval plan is unknown to Indian architecture.

Jantāghara: hot bath room, Mahāvaṁsa, XV, 3l, not in the Dictionary, though described without citation of the term, Indian Architecture, p. 14. S.B.E. XIII, p. 157, note 2. Cullavagga, V, 14, 3 and VIII, 8, 1; Mahāvagga, 1. 25, 12-13.

Kaḍaṅkara, Pāli kaḷiṅgarā: plank of a stairway, sopāna, Cullavagga, v, 21, 2.

Kalā: no reference to the kalās; see Venkatasubbiah, A., The Kalas, Madras, 1911, and do, with E. Mller, in J.R.A.S., 1914. The lists include such items as nagaramāṇam, vatthunivesam, dārukriyā, etc.

Kalābhara: artist, expert. According to the Gautama Dharma-sūtra, VI, 16, the kalābhara who is five years older than oneself should be greeted with respect as bhoḥ or bhavan. Haradatta explains kalābhara as one who lives by the kalās, i. the knowledge of music, painting, leaf-cutting and the like.

Kacuka: kacukaṁ... silāmayaṁ of Mahāvaṁsa, XXXIII, 25, is evidently rightly translated by Geiger as “a mantling made of stone” (for the Hhandhathupa) . This muat be the correct designation for the “casing” and “casing slabs” of archaeologists.

Kapota: should be translated “roll-cornice, “ “larmier.” It is the main cornice of a  building, derived from the edge of the thatch and the primitive drip-stone cut above cave dwellings to prevent the rain from running in. The synonyms of kapota, candra, lupā, gopāna, are significant; see candra-śālā. The rendering of kapota by “spout” should be avoided. As pālikā is abacus, kapota-pālikā should be a fillet above the kapota. Kern is undoubtedly right in rejecting the meaning “dove-cot,” so also in the case of viṭaṇka. Mṛcchakaṭika, I, 51 has kavālapa-viṭaṅka, glossed kapota-pālikā uparigṛha and translated in H.O.S. “dove-cot”; “dove-ridge” would be better. In reliefs, birds are commonly represented ae perched on roofs and mouldings. Utpala’s definition of kapota-pālikā quoted on p. 111 of the Dictionary, amounting to “corbel-ended timbers above the kapota” is quite intelligible, as these being seen end on, and coming between the top of the kapota, and the bottom of the next member above (as often represented in the early reliefs), are related to the kapota precisely as the abacus is related to the rest of the capital below it and the entablature above it.

Kappiya-bhūmi: not in the Dictionary. “Outhouse site,” Mahāvagga, VI, 33, 2=S.B.E., XVII, p. 119.

Karmāra, Pāli kammāra, Mahavagga 1, 48 etc., Sinhalese kammālar: not in the Dictionary. Artisan, smith, etc. Kammāra-bhaṇḍu, workers in metals, Mahāvagga, I, 48, 1. Highly esteemed by king and people, Jātaka, III, 281. The viceroy of Kṛṣṇarāya of Vijayanagar exempted kaṇmāḷars from taxation (A.S.I., A.R.., 1908-09, p. 184). Prakrit Kamāra, see Charpentier, Uttarādhyayanasūtram, p. 361. See also my Indian Craftsman, and Mediaeval Sinhalese Art. Kammmāra-sālā, smithy. p.261

Karṇa-kīla, “the ear rod, fastened with iron (nails), along the sides of a house, and according to which the house is to be built,” Arthaśāstra, III, 8. Probably the frame-work of four beams which rests on stone supports, cf. Mediaeval Sinhalese Art, Pl. VII, fig. 7, at the level of the man’s waist.

Kaṭaka: add, a position of the fingers used in  dancing, and seen in the hands of images holding flowers. See Rao, Elements of Hindu Iconography, I, p. 16; and Mirror of Gesture, p. 31. In this sense, synonymous with siṁhakarṇa.

Kaṭi-sūtra: in the sense of girdle, Cullavagga. V, 2, 1. Technical terms for special forms, ib. V, 29, 2.

Keyūra: armlet, cf. kāyura in Cullavagga, IV, 2, 1, S.B.E. XX, p. 69.

Khaṇḍa, door (the actual leaf or leaves), Arthaśastra, III, 8. Meyer makes it a single leaf. Shamasastry renders as equivalent to kavāṭa; the choice depends on the meaning assigned to aṇidvāra in the same passage. The door in any case would open inwards, hence Meyer’s rendering with reference to the obstruction of space between two houses cannot be quite correct.

Kicikkha-pāsāṇa: Māhavaṁsa, XXXIV, 69, stones apparently used as paving slabs round a stūpa, probably so called as being very smooth (cf. Skt. kijalka, filaments of a lotus). Childers gives the form kijakkha-pāsāṇa.

Kiṅkini-jālaya: network of bells adorning a vedikā, Mahāvaṁsa, XXVII, 16. Often seen on Bharhut and other early rail-copings.

Kīrti-vaktra: add synonymns kīrti-mukha, makara(i)-vaktra, makara-patra, siṁha-mukha; and Sinhalese kibihi, and kāla-makara of Dutch archaeologists. The inclusion of the term in the Mānasāra shows that the text cannot antedate the Gupta period, for the makara face as the crowning element of a toraṇa. is not developed before that time at the earliest, the crowning element in earlier types being plain or having the form of a triśūla or śrīvatsa.

Kośa-gṛha, store room, treasury: has triple underground cellar with many chambers, amongst which is a devatā-vidhāna, or chapel, with images of the Vāstu-devatā, Kubera, etc., Arthaśastra, II, 5.

Koṣṭhāgāra: a pair of storehouses are referred to by this name in the Sohgaura plaque inscription, and illustrated on the same plaque (Fleet, in JRAS, 1907). They are described as trigarbha, having three rooms; Fleet discusses this at length, but it is evident from the illustrations that these rooms are on three storeys, for the storehouses are represented as small three-storeyed pavilions; it is true that the roof of the top storey is “out of the picture,” but its supporting pillars can be clearly seeen. For another use of garbha as designating chambers of a many-storeyed building see under Prāsāda, the Lohapāsāda. See also prakoṣṭha, s.v. gṛha, dvāra-koṭṭhaka, and kuṇḍa.

Kūdu, see s.v. candra-śālā.

Kumbha (and kalaśa): I cannot see any evidence in the texts cited to justify the translation “cupola.” The jar in question has actually always the form of a jar, and is placed above the dome, cupola, spire, āmalaka, roof-ridge, or whatever otherwise forms the top of a building. Kumbha also=temples of an elephant, see s. v. hasti-nakha.

Kunḍa: a bowl used as a rain-gauge (vorṣamāna) and placed in front of s granary (koṣṭhāgāra) (Kauṭilya, Arthaśāstra, II, 5).

Kuṇḍikā: should be equated with kamaṇḍalu (not in the Dictionary) and explained as the water-pot carried by Brahmanical hermits and Buddhist monks, and provided with two openings, one a funnel at the side for filling, the other at the top of the neck, which is also the handle. Many examples have been found on Indian Buddhist monastic sites. The kuṇḍikā is carried only by deities of ascetic type especially Brahmā and Śiva, and by ṛṣis, and should not be confused with the amṛta-kalaśa, which has only one opening, and is carried by other deities, especially Indra and Maitreya. A full discussion of the Indian and Chinese forms by the present writer and F.S. Kershaw will appear in Artibus Asiae.

Kūtāgāra: regarding the kūṭāgāra-sālā in the Mahāli Sutta of the Digha Nikāya, Buddhaghosa, Sumaṅgala-Vilāsinī, p. 309, has the following, which I quote here from a letter received from Mre. Rhys Davids: “In that wood they established a Saṁgha-park. There, having joined the kaṇṇikā (ear-thing, corner of the upper storey) of the pillars (thambha, lit. supports) above by the saṁkhepa (holding together, fastening together) of the kūṭāgāra-sālā, they made the pāsada (terraced or balconied mansion) like to a mansion of devas. With reference to this the Saṁgha-park was known as the Kūṭāgāra-sālā.” Here, cf. saṁkhepa with kṣepaṇa in the sense of cornice; but I suspect a reference to brackets connecting pillars and kaṇṇikā (the Dictionary has karṇikā=upper part of the entablature); such brackets are very frequently represented in the errly reliefs (Bhsrhut and Sāñcī). Acharya’s Index has no entry under “bracket,” but there must have been a word or words in use for so common a structural feature.

Geiger’s “balconied windows” for kūṭāgāra, in Mahāvaṁsa, Ch. XXVII, is scarcely satisfactory; the pāsāda of nine storeys has 100 kūṭāgaras on each storey, and little paviliona, pajara or (candra)-śālā seem to be meant, such as are very common in Pallava architecture; e.g. at Māmallapuram, and cf. Jouveau-Dubreuil, Dravidian Architecture, fig. 4. The pavilion occupied by the Bodhisattva while in his mother’s womb is called a kūṭāgāra (Lalita Vistara, Ch. VII).

As Pāli paṇṇa-kuti and paṇṇa-sālā are synonymous designations of hermits huts, and as these are always single-storeyed cells, it follows that kūṭa-śālā need not be a room on the top of a building.

I am inclined to suppose that kūṭāgāra generally means simply “a house with a finial (or finials).” Cf. kūṭa, “finial” (vase) in inscriptions cited in Dict., p. 708. Gaha-kūṭa, Jātaka, I, 76. In Ceylon in the eighteenth century the use of such finials was permitted only in the case of devāles, vihāres, resthouses, and the houses of chiefs of Disāwa or higher rank. On this analogy the ultimate meaning of kūṭāgāra would be “honorable building.” In all the early reliefs, palaces, city gates, temples, etc., are duly provided with finials, while village houses lack them..

Kuṭi: not in the Dictionary as a separate word, but cf. gandha-kuṭi.

In the śūlagava (=Iśūnabali) ritual of the Gṛhya Sūtras (citations in Arbmann, Rudra, pp. 104 ff.) kuṭi=āyatana in the sense of shrines erected for Īśāna, Miḍhuṣī and Jayanta.

Under gandhakuṭi add: see full discussion in A.S.I., A.R., 1906-07, pp. 97-99, with mūlagandhakuṭi and śailagandhakuṭi cited  from Sārnāth inscriptions. Reference should also be made to the Sāñcī relief, north toraṇa, left pillar, front, second panel, showing the Jetavana, garden with the  Gandhakuṭi, Kosambakuṭi, and Karorikuṭi (Marshall, Guide to Sanchi, p. 58), “the three favourite residences of the Buddha.” Further references: Kern, Manual of Indian Buddhism, P. 28; Cunningham, A.S.I., Reports, XI, pp. 80 ff.; Sahni and Vogel, Sarnath Catalogue, P. 19, 211; Grünwedel, Buddhist Art in India, p.16.

In the Maṇimekhalai the small temple of Campāpatī, patron deity of Puhār, is called a guṭikā.

Kappiya-kuṭi, vacca-kuṭi, Cullavagga, VI, 4, 10.

Lepa: medium, glue, should be distinguished from sudhā, plaster. Vajralepa, “adamantine medium,” actually glue, see recipe in the śilparatna, Ch. 64 (my translation in Sir Ashutosh Mookerjee Memorial Volume); Mediaeval Sinhalese Art, pp. 118, ll9. Cf. Uttara Ramacarita, III, 40.

Sudhā-lepya, plaster and paint, Bodhgayā, 6^th-7^th century inscription, A.S.I., A.R., 1908-09, p. 154.

Likh: additional to the common meanings is that of “turning” (wood, etc.). S.B.E., XX, 78, note 3, is wrong in supposing that turning was unknown to ancient India. Metal, wood, and ivory are all turned at the present-day by means of hand-power devices quite unlike the European lathe (see Mediaeval Sinhalese Art, P1. VI, fig. 4, for ivory, and remarks ib. p. 141); turned stone pillars are highly characteristic of Cāḷukyan architecture (cf. Rea, Chalukyan Architecture,: p 5); and turning is certainly involved in the manufacture of many objects represented in early reliefs. It is significant that the Sinhalese name of the grooved spindle used in turning is liyana kanda, and the word liyana corresponds to likhitum used in Cullavagga, V, 8, 1 and V, 9, 2 with reference to turned wooden bowls and bowlrests. A meaning, “to turn wood, etc.” should therefore be given in Pali and Sanskrit dictionaries under likh. S.B.E., loc cit., trying to escape the meaning “turning” goes so far as to speak of using an adze on metal; a comical idea, if regarded from the standpoint of practical craft.

Another reference to turning will be found in the Mahāsatipaṭṭhāna Suttanta (D.N. II, 291=Dialogues, 2, p. 328), “even as a skilful turner (bhamakāra) “; the simile, (“drawing his string out at length,” etc.), implies the actually surviving Sinhalese technique.

Steatite boxes “turned on the lathe,” found at Bhīṭā and assigned to the eighth century B.C., are described in A.S.I., A.R., 1911-12, pp. 43, 93. For some other references to early turned objects see Rūpam, 32, pp. 122-123.

Liṅga: the following references are of interest in connection with the Deva-Rāja cult in Java and Cambodia: Simpson, in JRAS, 1888 cites numerous instances and regular practice of erecting lingams over the burial places of dead san̄yāsis. In A.S.I., Southern Circle, 1911-12, p. 5 “san̄yāsins are not cremated, but buried, linga shrines or brindāvana being raised to mark the spot.” Ib. 1915-16, p. 34, quoting S.I. Ep., 1914, “In the case of San̄yāsins ... a raised masonry platform is sometimes set up over the place of burial, on which a tulsi plant is grown, or a atone lingam is set up as though to proclaim to the world that the body buried below has attained to the sacred form of Śiva-linga.” E. Carpenter, Light from the East, being Letters... by the Hon. P. Arunachalam, 1927, p.63, quoting a letter from the latter regarding the tomb of his guru, “On the site where his body is interred is a lingam to which the worship is offered as to the Master.” For the Deva-Rāja cult and its supposed South Indian origin see F.D.K. Bosch, “Het Lingaheiligdom van Dinaja,” Tijdschr. T.L. en Volkenkunde, LIV, 1924.

Loha: is not iron, but brass or copper, bronze, etc.  I do not think that any example of an Indian Image made of iron could be cited, The roofing of the Lohapāsāda (Mahāvaṁsa, Ch. XXVII) was of copper or bronze. In Mahāvaṁsa, XXIX, 11, loha-paṭṭa is a sheet of copper used in the foundations of a stūpa, but we find ib. 12, ayo-jāla when an iron trellis is designated. One of the most important architectural references to loha is Mahendravarman I’s inscription at Maṇḍagapattu (Jouveau-Dubreuil, Conjeevaram Inscription of Mahendravarman I, Pondicherry, 1919); here brick, timber, loha, and mortar are mentioned as customary building materials. Copper nails are common finds on ancient sites. Other examples of loha will be found in the Dictionary under ābhāsa (!). Cf. also Siṁhalese pas-lo, an alloy of flve metals.

Loṣṭa: the use of loṣṭa, probably slag, in preparing a kiṭṭa-lekhanī, should be noted (Śilparatna, Ch. 64).

Makara-toraṇa: hardly an arch “marked” with a makara, but one springing from two makaras, and usually crowned by a full-faced makara or makarī.

Mañca: cf. taṅkita maca, stone couch, the altar of a yakkhacetiya, viz. the bhavana. of the Yakkha, Suciloma (Saṁyutta Nikāya, X, 3, P. T. S., ed. p. 207), glossed pāsāna-maca, thus synonymous with śilā-paṭṭa, see my Yakṣas, p. 20, note 3 (veyaḍḍi).

See also S.B.E., XX, 87, note 2, ib., 168, note 3; and 278, note 3; Mahāvaṁsa, XXVII, 39. Also Geiger, Māhavaṁsa, translation, p 204, note 3; the text has bodhiṁ ussīsakaṁ... sayanaṁ but this means the vajrāsana at the foot of the Bodhi tree (the description is of the Māradharṣaṇa), certainly not the Parinibbāṇa maca. Heṭṭhāmaca, Jātaka, 1, 197, probably the earthen bench outside a hut. Macaṭṭhāna, space for a couch, Culluvagga, VI, 11, 3 (Commentary). Cf. s.v. Paṭṭa, Sthāna and Vedikā. Re S.B.E., XX, 278, note 3, I see no reason why the paṭipādaka of a mañoa should not be fixed legs; no ancient representations or modern examples have trestles. The only trestles occur in connection with tables (hattha-pīṭha of Sumaṅgala Vilāsinī, II, 20, text 1, 163, and as seen on early reliefs) and modern daṇḍāsana (Mediaeval Sinhalese Art, P1. X, 1). Pīṭha of the Cullavagga may include both hattha piṭha and pāda, tables and footstools, hardly “chairs.”

The fact that mañca and piṭha were cleaned by beating does not prove that they were stuffed or upholstered: the actual support may have been made then as now of plaited cane or plaited webbing and anyone who has had experience of such beds will realise that they frequently need airing and beating.

Meru: reference should be given to E. B. Havell, The Himalayas in. Indian Art, and W. Foy, “Indische Kultbauten als Symbole des Gtterbergs,” Festschrift Ernst Windisch, 1914.

Nāga-bandha: is said to be a kind of window, and this would evidently be a perforated window with a design of entwined serpents; there are some in the early Cāḷukyan temples, and one more modern is illustrated in the Victoria and Albert Museum, List of Acquisitions, 1926, fig. 74. Cf. Siṁhalese nāga-dangaya. But nāga-bandha also means both in Ceylon and in southern India, the stop of a chamfer (Mediaeval Sinhalese Art, pp. 88, 129, and Jouveau-Dubreuil, Dravidian Architecture, pp. 10, 25, 42 and fig. 17); this stop often approximates in shape to a cobra’s hood. Cf. nāga, s.v. candra-śālā.

Nagara: add reference to the detailed description of a city in Milindapaha, V, 4 (also ib. I, 2 and II, 1, 9); the terms nagara-vaḍḍhaki, daḷha-gopura, gopur-aṭṭāla, koṭṭhaka, devaṭṭhāna occur. Another good description of a city is cited in Barnett, Antagaḍa Dasāo, p. 1, from the Aupapātika Sūtra.

Nāgara: the meaning “secular” as contrasted with satya, “sacred,” vaiṇika, “lyrical, “ and miśra, “mixed,” should be cited from the Viṣṇudharmottara, in relation to painting.

Nārāca,. etc.: the Dictionary has only “a road running east.” In the Sthānānga Stra6 we have vajja-risaha-naraya-saṅghayaṇe=vajraṛṣabha-nārāca-saṅghayaṇe, meaning “with joints firmly knit as if by mortise, collar, and pin.” Hoernle, Uvāsagadasāo cites Abhayadeva’s Sanskrit commentary, according to which vajja=kīlika, risaha=pariveṣṭana paṭṭa or encircling collar, nārāya=ubhayato-markaṭabandha. Or double tenon and mortise joint, and saṅghayaṇa=scarfjoint, five kinds being enumerated (for illustration of one see Mediaeval Sinhalese Art, fig. 75). One would have thought that vajja simply meant “firmly.” As regards pariveṣṭana paṭṭa cf. Mahāvagga, V, 11, “Now at that time the Vihāras were bound together by thongs of skin,” explained by Buddhaghosa (cited S.B.E., XVII, p. 31) as referring to the tying together of bhitti-daṇḍakādi “wall posts; eto.” This would seem to have been natural in the case of tho wattle and daub walls of the simple paṇṇa-sālās; but we do also find early pillars decorated with designs of interlacing ropes or thongs which may be vestigial ornament, and the roof of the shrine of tho Turbanrelic at Sāñci (south gate, left pillar, inner face) is bound by crossing ligatures which could only be described as pariveṣṭana paṭṭa. Atharva Veda, IX, 3 refers to the parts of a house that are knotted and tied (naddha) . A house (śālā) with grass sides has beams (vaṁśa), ties (nahana) and binding (prāṇāha), clamps (saṁdaṁśa) and “paladas” and “pariṣvajalaya.” See also Upamit.

Cf. Mediaeval Sinhalese Art, p. 114, “Nails were not used in ordinary building, but everything was fastened with rattans and other jungle ropes.” This refers to modern village practise.

Nayanonmīlana: p. 88 in Indian Architecture: my detailed account of the netra-maṅgalya ceremony should be cited, Mediaeval Sinhalese Art, p. 70 f.

Pāduka: should be cited also in the sense of sacred footprints, used ae a symbol (śrīpāda, Viṣṇupāda, etc.). The vacca-pāduka of a latrine are slso of interest, see S.B.E., XVII, p. 24; good examples have been found on monastery sites in Anurādhapura. Cf. vacca-kuṭi. Numerous lavatory sites are illustrated in Mem. A.S.C., Vol. 1.

Pālikā: should be translated “abacus,” with references to Tamil palagï Jouveau-Dubreuil, Dravidian Architecture, pp. 10, 25, 42, and fig. 17. See also kapota (-pālikā).

Pāṁśu: not in the Dictionary. Not translated where  it occurs as a permissible building materictl, Buddhsghosa, Comm. on Cullavagga, VI, 1, 2, cited S.B.E. XIII, 174; the other permitted materials being brick, stone, and wood. Paṁsu, taking all its uses into consideration, should here be rendered “laterite,” a common building material especially in Ceylon. In Mahāvaṁsa XXX, 7-9, where paṁsu is used in making bricks, the word is rendered “sand” by Geiger; but “decomposed rock,” “grit,” would be preferable. True sand (vālikā) would need only sifting, not crushing and grinding as well. In rendering such words some regard must be had both to practical considerations and to the materials actually available in a given locality. In the tropics the country rock decomposes either into true laterite (Sinhalese “cabook”) which is soft when cut, but hardens on exposure; or into a friable sandy grit; both of these have their use in building. Of course, there are many places where paṁsu means simply earth, dust, refuse, etc., cf. paṁsu-kūla, rags from dust-heap. See also śarkara, s.v. in Dict. and under ābhāsa.

Pacāṅgula: hattha-bhitti of Cullavagga, VI, 2, 7 explained by Buddhaghosa as pacaṅgula bhitti: pacaṅgulika-pantikā, Mahāvaṁsa, XXXII, 4; pancangulitale, Aupapātika Sūtra, §2. Possibly colored impressions of the human hand such as one not uncommonly sees on house walls, more likely a five-foliate design such as the palmettes which are so characteristic of early Indian decoration. In all the above passage we have to do with ornament applied to walls or to cloth. Cf. the “three-finger ornament” of Annandale, N., Plant and unimal designs...of an Uriya village, Mem. A. S. B., VIII, 4, fig. 2.

Pajara, which has, like candra-śāla-vātāyana, the double significance of ‘attic” and “dormer window” (see Jouveau-Dubreuil, passim), occurs in the latter sense in Jataka, III.379 “looking down from an open window (vaṭasīhapajarena) .” Cf. Mahāvaṁsa, XXVII, 16.

Ratha-pojara, the body of a carriage, Jātaka II, 172, IV, 60.

Parikhā: Mahāvaṁsa, XXV, 48 timahāparikha, “having a great triple moat.” See also under Harmya.

Paṭṭa: no reference to the meaning “frontlet, “ except that under vīrapaṭṭa we find “front-plate.” In the story of Udayana, Jacobi, Ausgewhlte Erzhlungen, p. 32, a sovaṇṇo paṭṭo is used to cover the brand on a manś forehead and is oontrasted with mauḍa, a turban or crown. In Ceylon the gold forehead plate used in investiturea is called a nalal-paṭa, those thus honored being known as paṭṭa-bendi. In Prabandhacintāmaṇi we get paṭṭa-hastin, state elephant; now elephants do not wear turbans, but do wear jewelled bands round the temples. In Bṛhatsaṁhitā the section on paṭṭas, which are not worn by those of the highest rank, seems to imply the meaning frontlet. Even Mahāvaṁsa, XXIII, 38, dukūlapaṭṭena veṭhayitvā may refer only to the tying on of a fillet, though “turban” seems plausible. No reference to paṭṭa in the sense of stone slab, etc. See Mālavikāgnimitra, III, 79 (silāpaṭṭāṁ) and Hoernle, Uvāsagadusāo, II, p. 107; sthala (Sthāla) as synonym, Mālavikāgnimitra, IV, 132. Loha-, and sajjhu- paṭṭa, sheets of copper and silver, Mahāvaṁsa,. XXIX, 11-12 Pāṭika, stone slab at the foot of the steps, Mahāvaṁsa, XXXI, 61; other terms current in Ceylon for “moonstones” are handa-kaḍa pahana (=candra-khaṇḍa pāśāṇa), and iri-handagala (=sūrya-candra kala). Ūdhva-paṭṭa, “stela,” should also be noted. Yogapaṭṭa is the braid used by hermits to support the knee when seated on the ground. Cullavagga, V, 11, paca-paṭika, perhaps a “cupboard with five shelves.” See also under nārāca.

Phailaka: commonly a panel for painting on. Add: appasena°, a board to lean against, when seated on a couch. to protect the walls, Cullavagga, VI, 20, 2, and VIII, 1, 4. Phalakattharasayana, a wooden bed, Jātaka, 1, 304. a kind of cloth, Mahāvagga, VIII, 28, 2 (see note in S.B.E., XVII, 246), and Cullavagga, V, 29, 3. See also s.v. Arghya and Pralamba.

Prākāra: an important reference is misplaced under prāsāda, Dictionary, p. 419. The Besnagar inscription (Mem, A.S.I., No. 4, pp. 128, 129) should be cited (pūjā-silā-pākāra); also Khāravela’s inscription at the Hāthigumphā, Udayagiri. The Mahāvaṁsa, XXV, 30, has ucca-pākāra, rampart; ib. XXXIII, 5, hatthi-pākāra, in the sense of the basement retaining wall of the platform of a stūpa with the foreparts of elephants projecting in relief (see also Parker, Ancient Ceylon, p. 284). Cullavagga, V, 14, 3 and elsewhere has itṭḥa-, silā-, and dāru-pākāras. Other references, Mysore A.S. Reports, 1913-14, pp. 8, 14 and 1919-20, pp. 2, 3, 5. In Kauṭilīya Arthaśāstra, 53, “rampart” rather than “parapets.” Pākāra=wall round a park, Buddhaghosa, Sumaṅgala Vilāsini, I, p. 41.

Pralamba, (-phalaka): reference should be made to the illustration of a pralamba-phalaka, fig. 94 in my Mediaeval Sinhalese Art, and the full explanation of its use there given according to the Sāriputra, as the Bimbamāna (see Dictionary, P.768) is called in Ceylon.

Pramāṇa: the single meaning given, “measurement of breadth” is insufficient. Promāṇa, in the sense of “ideal proportion” appropriate to various types is one of the ṣaḍaṅga of painting, given in Yaśodhara’s Commentary on the Kāmasūtra. See also Masson-Oursel, “Une connexion dans l’esth’etique et la philosophie de l’Inde, La notion de Pramāṇa,” Revue des arts asiatiques, II, 1925 (translated in Rūpam, No. 27/28). Pramāṇa=land area specified in grants, see Thakur in Sir Ashutosh Mookerjee Memorial Volume, 1928, p. 80.

Prāsāda: No reference to the Bharhut relief with inscription Vijayanta pāsāda, the only early prāsāda identified as such by a contemporary inscription; it is a three-storeyed palace (see HIIA, fig. 43); we possess so few positive identifications of this kind that none should be omitted. The Lohapāsāda described in Mahāvaṁsa, Ch. XXVII, was an uposatha house of nine storeys each with 100 kūṭāgāras “provided with vedikās, and it contained 1000 chambers (gabbha). It was covered with plates of copper, and thence came its name “ (ib. XXVII, 42); it was of wood, as it was later burnt down (ib. XXXIII), and rebuilt with only five storeys; the stone pillars on which the superstructure was erected are still standing at Anurādhapura. The Satmahal-pāsāda at Poḷonnāruva should also be mentioned (HIIA. fig. 287). See also under gṛha.

Puṇya-śālā,-gṛha: not in the Dictionary. Both have been thought to refer to temples, but the meaning dharmaśālā is far more probable, as pointed out by Hopkins, Epic Mythology, p. 71 (ib., 70-73 contains a very valuable discussion of images and temples as referred to in the Epics).

Raṅga, raṅga-bhūmi, nāṭya śālā, prekṣa-gṛha, etc.: not in the Dictionary. Nocitation in the Dictionary of theNāṭya-śāstra, where the construction oftheatres is described at some length, with muchuse of technical architectural terms. Araṅnga-bhūmi, stage, set up, Mahāvaṁsa,XXXI, 82. Raṅga, Jātaka II, 152.

Rathakāra: “car-maker,” carpenter, not in the Dictionary. A śudra, but connected with Vedic sacrifices; a snātaka, may accept food from one (Baudhāyana DhS., I, 3, 5=S.B.E., XIV, 159). Much information on the social position of craftsmen and related subjects is given in my Indias Craftsman, apparently unknown to the author: see also karmāra and āveṣaṇin, above, and rūpakāra, below. Rathakāra in inscription of Virūpākṣa I, A.S.I., A.R., Southern Circle, Epigraphy, 1915, p. 106.

Rūpakāra: sculptor, not in the Dictionary. But the Śilpin Rāmadeva, son of the rūpakāra Suhaka, inscription at Dhar, A.S.I., A.R., 1903-04, p. 240, is cited under Rāmadeva.  Reference should be given to Śivamitra, a śela-rūpakāra of Mathurā, mediaeval inscription at śrāvastī, A.S.I., A.R., 1908-09, p. 133. For Buddha-rakkhita, a rūpakāraka, see Cunningham, Bharhut, inscription No. 42.

Sabhā: the Bharhut relief with inscription Sudhammā Deva-sabhā, a pillared circular shrine with cornice and dome is not cited (HIIA, fig. 43). See also Seṁyutta, Nikāya, XI, 3, 5=Kindred Sayings, I, p. 307, and Dīgha Nikāya, II, 207-209.

In Jātaka VI, 127, the Sudhammā-sabhā of Indra has octagonal columns (aṭṭhamsa sukatā thambhā). The description of the heavenly sabhās in Mbh. II, 6-11, is altogether vague.

Sahasra-liṅga: not a “group” of a thousand phalli, but one lingam with a thousand facets, representing a thousand liṅgas. A good example at śrīśilam, A.S.I., Southern Circle, 1917-18, Pl. V.

Samudrāgāra: a summer house by a lake, Mālavikāgnimitra, Act IV. Samuddavihāra, a monastery on a river-bank, Mahāvaṁsa, XXXIV, 90. Samuddapaṇṇa-sālāya, ib. XIX, 26, a hall built on the sea-shore. Cf. the pavilions on the bund at Ajmer, and the island palaces at Udaipur.

Santhāgāra: “mote-hall,” with a central pillar (majjhima-tthambaṁ), Dīgha. Nikāya, III, 209=S. B. B., IV, 202.

śilpa: in the Atharva Veda, a “work of art” (Bloomfield, Atharva Veda, p. 70).

śilpa-śāstra: Hsan Tsang’s reference to five vidyās, of which the Śilpasthāna-Vidyā is one, is important as proving the existence of technical works on śilpa in his day (Beal, Records, I, p. 78). The much earlier Śulva Sūtras are effectively śilpa-śāstras, though not actually so designated.

śivikā-garbha, sivikā-gabbha: an inner room shaped  like a palankeen, Cullavagga, VI, 3, 3. Glossed by Buddhaghosa as caturassa, foursided. What may be meant may be gathered from the elaborate sivikās represented in Amarāvati reliefs, where their design is quite architectural (Burgess, Buddhist stupas of Amaravati and Jaggayyapeta, Pl. XI, 2 and p. 55, and Pl. XI, 1).

Sopāna: see s.v. ālamba-bāha, harmya, hasti-hasta, kaḍaṅkara, paṭṭa.

śreṇi: that painters were organised in guilds is apparent from Jacobi, Ausgawhlte Erzhlungen in Māhārāṣṭrī, P. 49, where the painter Cittan̄gaya, “working in the kingś citta-sabhā” belongs to a sreṇi of cittagaras. It is of interest that his daughter Kanyamajarī also paints. See also list of 18 guilds in Jātaka VI, 22: other references s.v. seṇi in P. T. S. Pali Dictionary.

śrīvatsa (sirivaccha) : also characteristic for Mahāvira. The cruciform flower is the later form only; in the Kuṣāna period it is what numismatists have called a nāga or shield symbol (good illustration on a coin, Rapson, Coins of the Āndhra Dynasty, pl. VIII, 207, reverse, and on Mahāvīra’s breast, Smith, Jaina Stupa of Mathurā, pl. XCI, right); the development of the early form into the later can be traced. Also cf. Hopkins, Epic Mythology, p. 205.

Sthāna: the sense of pose, stance, is not given. Five sthānas (frontal, three-quarter, profile, etc.) are defined in the śilparatna, Ch. 64, and thirteen in the Viṣṇudharmottara (see translation by S. Kramrisch, 26 edition, 1928). Mahāsthāna, sacred area, inscription of Mahīpāla. Saṁvat 1083, A.S.I., A.R., 1906-07, p. 99: Nāgendrasya.... Dadhikarnṇasya sthāne silapaṭṭo, Mathura inscription Luders’ List 85, Ep. Ind. I, 390, no. 18, cited Mem. A.S.I., Vol. 5.

Stūpa: no description of the component parts is given: they are sopāna, aṇḍa, medhi or garbha harmikā, yaṣṭi, chattrāvali, varṣa-sthāla or aṃṛta-kalaśa. There should be mention of the synonym dāgaba (dhātu-garbha), and of eḍūka and jāluka by which names Buddhist relic shrines are referred to in the Mahābhārata (3, 190, 65 and 67).  The detailed description of a stūpa in the Divyāvadāna, p. 244, summarised by Foucher’ L’Art gréco-bouddhique,.. I, p. 96, and the detailed account of the building of a stūps in Mahāvaṁsa, Chs. XXVIII, seq. should be referred to; also the full account in Parker, Ancient Ceylon. The letter quotes a Sanskritic-Pali text defining the shapes and proportions of dāgabas, from the Waiddyānta-pota (Or Vājayantaya) a śilpa-śāstra; well known in Ceylon, but not mentioned in the Dictionary. The Avadāna Śataka mentions three kinds of stūpas-gandhastūpa, keśanakhastūpa, and stūpa—the latter being the regular dhātu-stūpa for funerary relics. The Dhammapada Atthakathā, XXI, 1-290, H.O.S., Vol. 30, p. 175, has a thūpa built over the body of a Brahman’s son who had become a Buddhist monk. Were stūpas ever erected by others than Buddhists or Jainas? In Kāśyapa’s Conversion at Sāñcī (east gate, left pillar, inner face, third panel) a railed stūpa forms part of the Jaṭila ārāma: so also at Amarāvatī, Fergusson, Tree and Serpent Worship, P1. LXXXVI.

Stūpikā: cetiyasīse kirīṭaṁ viya kanakamayaṁ thūpikaṁ ca yojetvā (Attanaguluvaṁsa, Alwis, IX, 7). Dome of a palace, Mahāvaṁsa, XXXI, I3, with above reference (Geiger).

Cf. silāthūpaka, Mahāvaṁsa, XXXIII, 24, “a little stone stūpa,” probably actually the stūpa of II. I. I. A., fig. 292. But the usual meaning of stūpikā (as given in Dict.), is “dome.” I do not think this terminology implies a derivation of the dome from the stūpa, but only a resemblance of form. Granting the recognized resemblance, however, the point is of interest in connection with the origin of the bulbous dome, for many early stūpas are markedly bulbous. Some Pallava temples have bulbous domes, and even the dome of H.I.I.A. fig., ca. 200 A. D. almost exactly follows the shape of the slightly swelling aṇḍa of the stūpa of ib. fig. 146.

śulka-śālā: a toll-house, Divyāvadāna, 275, seq: śulka-sthāna, Arthaśāstra, II, 3.

Tāla-māna: here reference should be made to many published accounts, e.g.  Rao, Tālamāna, my Mediaeval Sinhalese Art, Ganguly, Orissa and her Remains. On pp. 230, 233, what part of the body is the “hiccough?“

Tṛṇacchadana, Pali tiṇa-cchadana: “thatch,” Cullavagga, passim. In Atharva Veda, IX, 10, 11, the thatch is called a thousand-eyed net stretched out like an opaśa on the parting (viṣuvant, here=ridgepole). See also Upamit.

Tulā: the meaning “well-sweep” should be added (Cullavagga, V, 16, 2); two other means of raising water are mentioned, loc. cit., viz. karakaṭaṅka literally “pot-edge“ or ”pot-ridge,” probably the “Persian” water-wheel, and cakkavaṭṭaka, wheel and axle. All three are still in common use.

But is karaka-ṭaṅka really distinct from kara-kaṭaka, a hand wheel for drawing water?

Upamit, etc.: RV. I, 59, 4 and IV, 5, 1; AV, IX, 3, 1. See Bloomfield, Atharva Veda, II, 185, 195; Whitney, Atharva Veda, 525; Zimmer, Altindisches Leben, Ch. V; etc.

The whole terminology of the śālā is difficult, but the rendering of upamit as (sloping) buttress (by Bloomfield and by Zimmer) is extremely implausible and almost certainly an error. I suggest upamit=plinth or pillar base; such bases were probably, as at the present day, of stone, as a, protection against white ants.7 Then pratimit (=sthūṇa) are the main upright wooden pillars (corner pillars) set up on the upamit; parimit, the horizontal beams of the framework, connecting with the pratimit by means of mortices or dovetails (saṁdaṁśa);8 pakṣa, perhaps the wall plates; vaṁśa, the bamboo rafters. The roof (chanda) is thatched with straw or reeds (tṛṇa): the cut ends of the reeds may have given rise to the designation “thousand eyed” of AV. IX, 3, 8. Palada (bundles of grass or reeds, according to Zimmer) and pariṣvajalaya I cannot explain.

The śikyāni, ropes “tied within for enjoyment,” may have served as partitions, to be hung with cloths so as to divide the interior into separate rooms; the Sinhalese pilivẹla, is used in this way, and I remember to have seen an ornamental example carried by a party of travellers for use in a public resthouse to secure privacy.

Vajrāsana: “diamond throne,” though well-established, not a good rendering; “adamantine throne” would be better. See E. Senart, “Vajrapāni dans les sculptures du Gandhara, “ Congr. Int. Orientalistes, Alger, 1905, Vol. I, p. 129. Bodhi-pallaṁka in the Nidānakathā, Jātaka, I, 75, is an interesting synonym. The Buddha’s āsana at the Gal Vihāre, Poḷon̄āruva, Ceylon, is decorated with actual vajras, hut this probably represents a late interpretation of the term; I know no other instance. See also Bodhi-maṇḍa and Mañca.

Vāna-laṭhī, rafters or reepers? As a protection  against the rain, the vānalaṭhī (of a house, gṛha) are to be covered over with straw (kaṭa, here thatch rather than straw mats), Arthaśāstra, III, 8. Cf. Yaṭṭhīvana.

Vapra: in Kauṭilīya Aṛthaśāstra, 51, 52, vaprasyopari prākāraṁ; “glacis” rather than “rampart,” which latter rises above the vapra.

Vardhaki: I cannot think of any case where the vardhaki, Pali vaḍḍhaki, is specifically a painter. The usual meaning is architect, artisan. Cf. nagara-vaḍḍhaki, the architect of a city, Milindapaha, II, 1, 9. In Mahāvaṁsa, XXX, 5, the 500 iṭṭhakā-vaḍḍhakī are certainly not all “master-builders” as rendered by Geiger, but rather brickmakers or bricklayers; even the vaḍḍhaki who is their spokesman, ib., 12 is hardly more than primus inter pares. Vaḍḍhaī, architect, one of the 14 ‘jewels’ of a Cakravartin, Uttarādhyayanasūtra commentary, cited Charpentier, p. 321. Numerous designations of craftsmen will be found in the Śatapatha Brāhmaṇa list of symbolic victims of the Puruṣamedha (S.B.E., XLIV, 413-417).

Vardhamāna: add “powder-box,” one of the aṣṭamaṅgala of the Jains. Early illustrations, Smith, Jain. Stupa of Mathura, pl. VII; later, Httemann, “Miniaturen zum Jinacarita,” Baessler Archiv., 1913, fig. 1. Vardhamāna-gṛha, Uttarādhyayanasūtra, IX, 24.

Vastra-nip(y)a: is not “a jar-shaped ornament of a column,” but the knotted band or ribbon which so often encircles the pūrṇa-kumbha which forms the base or capital of a column, and the Mānasāra text cited (kumbha-madhye, etc.) is perfectly explicit on this point, “and in the middle of the pot (i.e. round the belly) let there be added a colored band of cloth as a protection.” This use of a string or band as protecting charm or “fence” is of course well known in many other connections.

Vāstu, add the meaning “real estate” (Meyer, “Liegenschaft”) : “Vāstu includes houses, fields, groves, bridges (or ghāṭs, setu-bandha), ponds, and reservoirs, “ Arthaśāstra, III, 8.

Vātāyana: the Dictionary citations show that in the śilpa-śāstras types of vātāyana are differentiated by preceding qualifying adjectives denoting the pattern of the grille or openwork screen. In the light of this fact, and of the varieties of windows represented in reliefs and the types still in common use, the three designations in Cullavagga, VI, 2, 2 are perfectly intelligible: vedikā vātapāna is a window with a rail-pattern grille; jāla-vātapāna is one with a trellis grille, lattice; salāka vātapāna, one provided with upright turned pillars or bars (not “ slips of wood”) . Buddhaghosa glosses salāka as thambaka. For turning, s.v. likh.

Vedī, vedikā, etc.: veiyā of Jacobi, Ausgewhlte Erzahlungen, p. 49, must be marriage pavilion rather than balcony, as marriages always take place in special temporary pavilions erected ad hoc.

In the common sense of railing, the Mahāsudassana Sutta, I, 60, gives the component parts, viz. stambha, (uprights), sūci (cross-bar), uṣṇīṣa (coping), and these words often occur in Prakrit forms in the early inscriptions: also plinth, ālambana. In Mahāvaṁsa, XXXV,2, muddhavedī is the railing of the harmikā, pādavedī the railing on the basement level of a stūpa; ib. XXXVI, 52 and 103 has pāsāṇa- and silā-vedī, “stone railing” (round the Bodhi-tree) rather than “stone terrace” as interpreted by Geiger, p. 296.

Māhavaṁsa, XXXII, 4, vedikā represented in a painting. Ālambabāha, the vedikā of a, sopāna, Cullavagga, V, 11, 6 etc. See also kiṅkini-jālaya. Cross references to p(r) ākūra, and bhitti, should be given; cf. bhitti-vedikā of Mālavikāgnimitra, V, 1, where it is built round an aśoka tree.

The very curious use of vedikā to mean a mode of sitting (āsana) is noted by Charpentier, Uttarādhyayanasūtram, p. 371.

Vidyut-latā: Pali, vijjul-latā, Mahāvaṁsa, XXX, 96, the Commentary having megha-latā nāma vijju-kumāriyo, “the cloud-vines called lightning maidens.” Real lightnings are evidently intended, not mere zigzag lines as rendered by Geiger. Representations of clouds and lightning are very characteristic of Indian painting; certain rooms in the old palace at Bikanir, entirely decorated with a frieze of clouds, lightning, and falling rain may be cited (see my Rajput Painting, P1. VII). The form vijju-kumāriyo is interesting, as the lightning is similarly always feminine in relation to clouds in rhetoric, and cf. Yajur Veda, IV, 1, 11, Jātaka, V, .407 and Mṛcchakaṭika, V, 46.

Vimāna: reference should be made to the long and excellent discussion of this word in the P.T.S. Pali Dictionary.

Vīṇā: as this word and also karuṇa-vīṇā are separately rendered “flute,” there can hardly be a misprint; the proper word is, of course, lute. Two forms are found in the early reliefs, one like a harp, the other like a Japanese biwa. So far as I know the southern vīṇā with two large gourds as sounding boxes can be seen first in the paintings at Elūra. The parts of a vīṇā are named in Milindapatha, II, 3, 5; see also P.T.S. Pali Dictionary s.v.

Historical Architects, add:

Ānanda, son of Vāsiṣṭhī, as above, s.v. āveṣaṇin.

Balaka, pupil of Kaṇha, maker of a śālikā at Kondae, and one of the earliest craftsmen known to us by name (Burgess, Report on the Buddhist Cave Temples, 1883, p. 9).

Bammoja, western Cāḷukya inscription. Bammoja was “a clever architect of the Kali age; the master of the 64 arts and sciences; clever builder of the 64 varieties of mansions, and the inventor (?) of the four types of buildings called Nāgara, Kāliṅga, Drāviḍa, and Vesara” (A.S.I., A.R., 1914-15, Pt. I, p. 29), The description of Kāliṅga as a style is cited in the Dictionary from the Mānasāra.

Dīpā, builder of the Caumukh temple at Rāṇpur; belonged to the Sompura class of Brahman architects, whose ancestor is said to have built the temple of Somnāth-Mahādeva at Prabhās-Paṭṭan. The Sompuras, not mentioned in the Dictionary, are said to have built many temples in Gujarat, to have been at Ābu, and to possess MSS. on architecture. One, Nan̄ā-khumma, was in charge of repairs at Rāṇpur; another, Keval-Rām constructed temples at Ahor (D. R. Bhandarkar, “Chaumukh Temple at Rāṇpur, “ A.S.I., A.R., 1907-08).

Jaita, etc.: an inscription on the window of the second storey of Rāṇa Kumbha’s kīrtistambha at Chitor (A. D. 1440-49) mentions the architect of the building, and his two sons Naps and Puja. On the fifth storey are effigies of the two last, and a third son, Pama.

Another inscription at Chitor mentions the fourth son, Balrāja. See A.S.I., A.R., 1920-21, p. 34.

Sidatha (Siddhārtha), son of Nāgacana, as above, s.v. āveṣaṇin.

Śivamitra, as above, s.v. rūpakāra.

Mallikārjuna Chin̄appa, builder of the Vīrabhadra temple at Chikkabaḷḷāpur, Mysore, died 1860; there is a tomb (gaddige) in a building to right of the temple.

Treatises on architecture:

Bimbamāna: known in Ceylon as Sāriputra. Add reference to translated passages in my Mediaeval Sinhalese Art.

• 1. ? yal lekhyaṁ
• 2. See S.B.E. XX, p. 105, note 2.
• 3. Pāli pakuṭṭa, Cullavagga VI, 3, 5 is rendered “inner verandahs” in S.B.E., XX, p. 175.
• 4. But see Parikhā, usually, and perhaps here also, a moat.
• 5. Or, if we read asaṁhāryo, then supporting a fixed bridge.
• 6. Benares edition, p. 413a, cited by Hoernle, Uvāsagadasāo, II, Appendix, p. 45.
• 7. Cf. Mediaeval Sinhalese Art, p. 129, fig. 72, and pl. VII, fig. 7, “Wooden pillars often rest on a stone base as a protection against white ants.”
• 8. Mediaeval Sinhalese Art, loc. cit. (p. 129), “where the whole building rests on low stone pillars, the wood pillars are mortised into huge beams forming the framework of the floor.”

         Vedic parimit and Sanskrit karṇa-kīla seem to designate such foundation beams; Vedic pakṣa and Sanskrit karṇikā the wall plates forming the framework of the roof. Where we have to do with a colonnade rather than a wall, karṇikā is of course ‘entablature.’