1871.] 
The Rock-cut Excavations at Harchoka. 
179 
The section drawing represents the temple reconstructed, for it 
is now almost entirely destroyed, and I doubt if in another fifty 
years any traces of it will be left. The river just above the temple 
takes a very sharp bend, in the re-entering angle of which a creek 
has been formed by the eroding action of the water, (Vide PL vi.) 
and from the bank of the river being in this part low and shelving, 
the river, during the temporary impetuosity of floods, dashes with 
full velocity into this creek, and sweeping over the bank completely 
submerges the temple, which, at the highest floods is, throughout 
its greater portion, about two feet below the surface of the water. 
It was a work of considerable time and labour to clear the ruins 
of the debris that lay within, but when this had been done, it was 
satisfactory to find that enough remained to enable me to form a 
yery correct idea of what the excavations had been like in their 
perfect state. As a whole they assume rather an odd form, but this 
must be attributed to the limited space that was available for the pur¬ 
pose, the cell nearest the river being only nine yards distant from it. 
I am inclined to think that these excavations have been made at 
three distinct periods, and for this reason I have, for distinction’s 
sake, styled the three parts the upper, middle, and lower temples 
respectively. The lower has all the appearances of being the 
oldest. None of the columns in that part remain intact, but from 
the lower portions that remain, they appear to have had no base 
like those in the hall of the middle temple. The latter is, in my 
opinion, the latest excavation of the three. The two columns which 
remain standing in the upper temple are very rudely fashioned. 
The middle temple on the other hand has a much more finished 
appearance, the walls of the hall being rubbed quite smooth. 
I could find no inscriptions in the lower temple. The inscription 
marked B was found on the verandah-wall of the upper temple, 
near the entrance to one of the cells. The inscriptions marked 
A, C, D, E, E, and Gr were cut on the hall pillars, only five of which 
I found standing. 
The doorway shown in the section drawing, in which a pair of 
grotesque looking figures* support the scroll of lotus, stem and 
flowers, is the only attempt at ornamentation that is to be seen. The 
other doorways are each set off with a border of plain mouldings. 
# These figures were called Kirttimukh by the Mukarrir mentioned on p. 180. 
