KAU.i INDIA. 35 



liardly would recognize what they are intended to recall ; and it may be 

 added that no obscene conceptions are mingled in the minds of the many 

 thousands of Hindoos who venerate imder this form the generative energy 

 of nature. The great centre of Siva-worship in India is the city of Benares. 

 After this digression, I insert Mr. Rivett-Carnac's description of the Chan- 

 deshwar temple: — 



"On visiting the temple sacred to Mahadeo at the entrance to the 

 gorge, I could not help being struck by the peculiar construction of many 

 of its shrines as bearing a marked resemblance to these rock-markings. In 

 addition to the principal shrine, placed within the temple itself, a massive 

 little structure built up of large stones, many of which would appear to 

 have been taken from Buddhist ruins so plentiful in the neighborhood of 

 Dwara-Hath, I counted thirty-seven minor shrines within the walled in- 

 closure by which the temple is surrounded. These consist mostly of a 

 rough pedestal formed of loose stones surmounted by a Mahadeo and Yoni. 

 The Yoni, in the largest of these shrines, was a solid block of stone, cut to 

 the well-known 'jew's-harp' shape, the upright Mahadeo being slightly 

 carved at the summit and base. Some half a dozen others were more or 

 less solid and well made, according to the conventional construction of these 

 symbols. In one case the stone which did service for the Yoni was the 

 cushion-shaped finial of some Buddhist temple, the Mahadeo being repre- 

 sented by a carved head with higli-raised cap, broken off from some neigh- 

 boring ruin. The fragment had been inserted, cap downward, in the square 

 hole by which the cushion had been fixed on to the top of the original 

 structure." 



I interrupt here the author's account in order to direct attention to 

 Figures 27 and 28, the first of which, copied from Plate III of the pamphlet 

 under notice, represents the section of a large stone Mahadeo and Yoni in 

 the Chandeshwar temple ; while Fig. 28 shows the same symbol in a more 

 elaborate form, as seen by the author in a temple or shrine at Benares, and 

 illustrates the "jew's-harp" shape to which he alludes. In this instance, 

 by way of attribute, a serpent is coiled around the emblem of Mahadeo. 

 The figure is taken from another pamphlet by Mr. Rivett-Carnac, relating 

 to the snake symbol in India. Leaving aside the serpent, a ground-plan of 



