The bsTan ’gyur includes four Indian works specifically devoted to iconometry, but, in practice, the theory of the proportions of the image of the Buddha in Tibet is based upon three religious texts found in the Buddhist canon: chapter 30 of the Mahāsaṃvarodaya Tantra; chapter 5 of the Kālacakra Tantra; and the Pratimālakṣaṇa Sūtra, generally known in Tibet as Sha ri’i bus zhus pa’i mdo, of which four different versions have been known to Tibetan artists (Tucci, 1980: 291-292). The Mahāsaṃvarodaya Tantra states that the figure of the standing Buddha measures 120 digits (aṅgula, sor mo). However, the Kālacakra Tantra asserts that the figure of the Buddha measures a few more digits than 120, which led commentators to interpret this as meaning 125 digits (Peterson: 241-242, 246, table I; Jackson: 144-147). This prompted Ratnarakṣita, one of the last Indian scholars to find shelter in the Nepal Valley in the first half of the thirteenth century and the author of a commentary on the Mahāsaṃvarodaya Tantra, to amend the measurements of the Buddha figure contained therein, by stating that they amounted to 125 digits (Jackson: 145, 147, n. 14), thus implying that the Tantra was in error and should conform with the Kālacakra Tantra tradition. The lack of clarity of the Kālacakra Tantra on this point aroused discussions which lasted for centuries, but generally the theory of the five extra digits prevailed, being accepted by the great scholar and painter Padma dkar po (1526-1592) (BKNS: 310) and by other[page 474] artists down to this century. One of the greatest scholars in Tibet, the regent Sangs rgyas rgya mtsho (1653-1705), wrote a text recording the proportions of the eight different types of stūpa (Tucci, 1980: 136-137) and attempting to solve the discrepancy between the Kālacakra Tantra and the Mahāsaṃvarodaya Tantra. In that text, which is part of the Vaiḍūrya g.ya sel, an encyclopedic work devoted to astrology, chronology and history, Sangs rgyas rgya mtsho tried to reconcile the two traditions by suggesting that the measurement of 120 digits ought to apply to painted figures and that of 125 to statues, thus allowing for the additional depth of three-dimensional images (Jackson: 144; Peterson: 243). The regent’s suggestion was rejected by the famous artist and scholar Zhu chen Tshul khrims rin chen (1697-1769), who found out that it was the consequence of a spurious interlinear note added by a scribe or editor in a treatise written by the religious artist ’Phreng kha ba dPal blo bzang po (1543-1588) (Jackson: 145). Tshul khrims rin chen followed the tradition of allowing 125 digits to the Buddha figure, 120 to peaceful bodhisattvas, and so forth, with a decreasing number of digits for each of the four other categories of figures, according to the sixfold classification he adopted (Jackson: 50).

Tibetan scholars could not agree on the number of the categories of figures either. Bu ston and the eighth Black Hat Karma pa Mi bskyod rdo rje (1507-1554), for example, apparently divided them into eleven groups, but the Tibetan encyclopaedist Klong rdol bla ma (1719-1805) reduced them to four (Jackson: 50, 67, n. 4; Tucci, 1980: 299) and adopted the 120-digit measurement for the Buddha figure. Present-day artists, such as the eastern Tibetan painter dGe ’dun (18-19), are aware of the co-existence of two different traditions. Furthermore, later Tibetan writers on iconometry pointed out that several categories of figures could not be traced to canonical sources. In particular, Rong tha Blo bzang dam chos rgya mtsho (1863-1917) states that the measurements of the proportions of two classes of wrathful deities originated from an oral tradition which was based upon the correct measures of ancient Indian images (Jackson: 147).

Since Tibetan artists in practice resort to detailed drawing displaying the proportions of the various categories of figures rather than to textual sources (Lo Bue, 1990: 188-194), it may be suggested that iconometric texts are seldom more than displays of erudition by literati who are little concerned with their practical application[page 475] by artists. They are often incomplete and tend to take for granted a great deal of knowledge from the reader, in a manner that is characteristic of their Indian models. Padma dkar po (BKNS: 309, 312), for example, is very helpful when he explains the Indian numerical symbols used in the Śrīcaturpīṭha Tantra, where the “eyes of the sky” means 20 digits; the “king,” 16; the “sun,” 12; and the “water treasure,” namely the ocean, 4. However, his description of the measurements of the proportions of the stūpa of the Enlightenment type leaves out those of the discs making up the spire, and suffers even when compared with Klong rdol bla ma’s description, however incomplete (Padma dkar po, CTGK: 319-322; Klong rdol bla ma, ZDSB: 760-761). This kind of carelessness is typical of a scholarly literature that is chiefly aimed at the accumulation of religious merit rather than at the transmission of practical information.