of the temple on the Takht-i-Suleiman hill at Srinagar
are as old as the second century B.C., and that the
little rock temple that I visited beyond Bawan at
Bhumju belongs to a rather later period, but this is
far from probable, and the majority reduce their
age considerably. The greater part of the ruins
now to be seen was probably erected between
the fifth and tenth centuries. Sir Alexander Cunningham
places, the reign of Ranaditya, the builder
of Martand, between 370-500 a . d . , but Fergusson does
not admit its foundation before the eighth century.
Another difficult question is that of “ roofs,” for it is
almost impossible now to determine whether the great
fanes at Martand and Avantipura were covered with
slabs of the same beautiful blue limestone of which they
are built, or whether they were merely roofed with wood,
or as some have suggested, were left open to the heavens
—an improbability, one would imagine, in a climate
liable to long periods of rain and snow. If a foreign
influence is admitted—and it is almost impossible to
deny it—it was probably wielded by the Greeks who
settled in Afghanistan and the Punjab frontier under
Euthedymus as early as the third century B.C.
Besides the more extensive remains, there are small
temples in good preservation at Ladoo, on the right
bank of the Jhelum, a few miles below Avantipura, at
Payech (this one is extremely perfect, the capitals of
unusual beauty and grace), at Narastan, and in less
good preservation at many other places. The miniature
size of many of these is a curious feature. Excepting
the great temples at Martand and Avantipura, very
few attain to larger dimensions than those of a “ down-
country ” shrine. The details are always carried out
with loving care, and the workmanship is very perfect,
so that there can have been no idea of saving of labour.
Possibly the havoc wrought by the great natural forces
—by earthquake, floods, and fierce storms—filled with
fear the pious builders, and, lovingly careful of their
handiwork, they preferred a form better fitted by reason
of its compact solidity to withstand such dangers, to
one more imposing in size but less likely to endure.
The fear of invasion by folks of other creeds may also
have encouraged the building of less noticeable fanes.
Certainly the smaller ones have escaped to a large degree
the destruction wrought by iconoclastic rulers on the
greater structures. The tiny temple models, sometimes
found on pillars, are generally supposed to be funeral
monuments.
Buddhism has left very few traces in the land. A
rare stone with an image—possibly a Buddha—is discovered
from time to time; otherwise the influence of
those gentle teachers seems to have been swept entirely
away. Far more living is the belief in the “ Naga”
(snake-worship). Half their religious rites are connected
with this superstition, and few of the temples are
built without either a moat round or a tank in front in
honour of this god. I t is, perhaps, in recognition of the
faithful devotion offered that he keeps his myrmidons
from worrying the inhabitants too much, poisonous,
or certainly fatal bites, being small in number. The
temple of Pandrethan has one of the largest moats
I saw. This is a very perfect specimen of the later
Kashmir style; ornament had begun to run riot, and
the duplication of the parts betokens the change that
was coming over the builders’ art. Cunningham assigns
as late a date as 930 a . d . to it, believing it to have been