. astrology and the worship of the planets 241
The Tejaprabha cult enjoyed considerable popularity and was fairly
widespread throughout the Tang empire. Related icons have been
found as far apart as Chang’an^50 in Shazhou (Dunhuang)^51 and
in Sichuan, where the only extant examples in stone have been found
(figure 1).^52 At some point, evidently toward the end of the Tang, there
is an indication that the iconography relating to Bhaisajyaguru, espe-̣
cially in its astral aspect, was amalgamated with that of Tejaprabha.^53
The cult of Tejaprabha was of enduring importance up to the end of
the Yuan dynasty in China. It quickly spread beyond China’s borders
throughout East Asia and by the beginning of the eleventh century
it was equally present in the Liao and Koryŏ kingdoms as well as in
Heian Japan.^54 In both Korean and Japanese culture, the Tejaprabha
cult has continued well into the premodern period as documented in
votive paintings dating from the seventeenth to nineteenth centuries.
Astral Mandalas
The worship of the astral deities eventually found its way into the heart
of the ritual complex of Tang Esoteric Buddhism, in which the use
of mandalas is a central feature. Hence, a number of astral charts or
“astral mandalas,”^55 throught to have originated in the context of Eso-
teric Buddhism of the second half of the Tang, have been preserved
in Japan.
Among these are mandalas depicting the divinities of the Great
Dipper, the twenty-eight constellations, and the Western zodiac
(^50) Now in the collection of Geijutsu University in Tokyo. This sketch is dated 1148
C.E. and is considered a copy of an original from the Tang. Whatever the actual date,
the iconography is consistent with the Tejaprabha paintings from the Tang as found
in Dunhuang. Cf. Takeda 1995, 9–11. The author refers to this sketch as a “Great Dip-
per Maṇḍala,” but this is of course incorrect. The Tang depictions of Tejaprabha and
his astral retinue have nothing to do, textually or ritually, with a mandala per se. 51
Cf. Meng 1996. A good example dated from 897 C.E. is presently in the collection
of the British Museum. Cf. Whitfield 1982–1983, 323–24, pl. 27. 52
See Segalen, de Voisins, and Lartigue 1924, pl. CXXX.
(^53) Line drawings and sketches from the late Heian period provide us with evidence
of this. Cf. Hayashi 1997, 37–40.
(^54) For the cult of Tejaprabha under the Koryŏ, see Sørensen 1995c.
(^55) The designation “mandala” to these charts is due to the fact that they resemble
traditional Zhenyan/Shingon mandalas in their arrangement. However, in ritual terms
the use of these items do not quite follow the same procedure. A term such as “chart”
or “diagram” is probably more appropriate for these tableaux of astral divinities. This
problem was first pointed out by Howard 1983, 6–8, n. 8.