Page images
PDF
EPUB

surface into seven well-defined vertical panels. These seven panels are counterbalanced by the seven storeys of the tower, these seven storeys being brought about by five actual rows of miniatures making up five storeys, with one miniature on each face of the square tower to compose the sixth, and the actual summit of the tower itself constituting the last or seventh storey. For the rythm which I postulate it is obviously uncssential that there should be precisely as many rows of miniatures as there are vertical panels. What is essential is that the tower should show as many horizontal bands as there are vertical panels. But once constituted, these horizontal bands may be decorated or otherwise treated in a variety of ways, according to the aesthetic feeling of the architect. Thus we find that in the case of the Rām-ji Mandir at Semaria in Saran (Plate 8), the side of the cella wall is divided into nine vertical panels, and that the tower above the eaves is divided correspondingly into nine horizontal bands, some of these bands being horizontal rows of miniatures, and some being treated otherwise. The builder fortunately recognized that a mere multiplication of rows of miniature sikharas would result in intolerable monotony, and he has done his best to overcome this difficulty.

The culmination of the principles we have been following is reached in the Ramchandra Mandir at Ahalya Asthān at Ahiāri in the Darbhanga District (Plate 9) which comes perhaps nearest of the Tirhut temples to that type of modern temple in Benares which Fergusson illustrates. At first sight it is not altogether obvious how this peculiar building is to be accounted for, nor how it falls into line with all the types we have just seen. It is, however, apparent, that as regards the panelling on the cellawall, the Ramchandra Mandir is but one step in advance of the Ram-ji Mandir at Semaria, which was shown in the preceding plate. That temple had nine panels; the present one shows eleven; although in this ultra-developed form they are no longer actual projections from the sides of the cella, but purely decorative panels. Their number, however, is eleven, and we feel instinctively that, in the light of all the other temples of this general

class which we have examined, this number of eleven should be rythmically balanced in the storeys of the tower. Now to erect anything like an eleven-storeyed-tower on so small a base or cella as we see in this example would be a difficult and certainly most unæsthetic proceeding. The building is top-heavy even as we see it. The builder, therefore, was faced with two alternatives, as I conceive it, either to abandon the fundamental principle of construction, or to interpret it to fit his structure. He wisely chose the latter of these alternatives, and contrived to retain his rythm and keep his eleven storeys in a most ingenious way. At either side of the tower we see that he has built up a series of five rows of miniatures, putting five such in the bottom row, four in the second, three in the third, and so on, with the result that these series rise in a definitely ascending scale. The eye travels up them. The sixth storey, however, which crosses the entire width of the pinnacle, serves as a transitional member, and brings us to the top of what is as clearly a descending series of single miniatures centrally placed and so constructed that the topmost one is the largest, and the others each smaller and smaller, so that here the progress of the eye is inevitably downwards. Thus, having ascended by the rising tiers at either side, and having thus accomplished five storeys, the eye of the beholder meets the transitional sixth storey at the top, and is then led downwards through yet five other storeys, in this ingenious way completing the number of eleven requisite for rythm. Thus, although at first sight this temple at Ahiāri appears a mere monstrosity quite as inscrutable as Fergusson's temple in Benares we can now see that in reality it constitutes the legitimate culmination of the style. There are faint traces of a curvilinear outline here as in the Benares example, but to my mind these are accidental, and in no way essential. They are sufficiently accounted for by the long familiarity of the Indian eye to the curvilinear outline of the Orissan type of shrine, but I cannot see that this curvature has any fundamental bearing on the problem of development in the type of temple in the Tirhut region. The class as a whole appears to me to be essentially distinct from the

Indo-Aryan type of Fergusson's discussion, and to be explained as we have seen on very simple lines in perfect independence of the Orissan style, save perhaps in individual instances where the outline of the sikhara has been influenced in a minor way and to an almost negligible extent.

Now it will have been noticed that in all the monuments that we have so far seen, the tower is square in plan. A very simple development from this original norm was effected by cutting off the four corners of the tower, with the result seen in the Shiva Mandir at Dandaspur in Saran District (Plate II) where we see an octagonal tower decorated vertically as in the first group illustrated at the beginning of the lecture. The same principle is again illustrated by the Rādhā Krishna temple in Muzaffarpur (Plate 12) where an octagonal tower is treated horizontally. Here the projections on the side of the cella being two in number, the panels number five and the horizontal banding of the sikhara is measured or spaced with reference to this fact, although the pinnacle is left to do duty for the two uppermost storeys. In Plate 13, the Shiva Mandir of Ganpat Rām at Bagahā in Champaran, we have a similar octagonal tower with seven horizontal storeys, if we count the miniatures immediately above the entrance; but whether the side of the cella wall shows the three projections which this number of storeys indicates is unhappily not determinable from the photograph before us. In these more developed forms we sometimes find that the rythm has been overlooked, sometimes, but by no means always. Some octagonal towers, however, appear to show an even number of storeys, which could never be the case were the fundamental principles borne in mind. Thus the miniatures piled at the angles here appear to mount through six storeys only; but the appearance is misleading, the tower being as we have seen seven-storeyed in reality. The same may be true elsewhere also where the rythm seems at first to have been overlooked.

From the octagon the plan of the tower passed by a natural transition into the true circle, and we get the temple with simple round tower, such as is shown by the Bhagawati Mandir at

Subegarh in the district of Muzaffarpur (Plate 15). This building I am told was erected by an exile from Nepal, and forms perhaps no very genuine link in the chain of development in Tirhut. I venture to illustrate it here, however, as it fills what would otherwise be a gap in our logical series. There appears to be no other example of this type in the whole division, which perhaps is not regrettable on æsthetic grounds.

From the tower we pass next to the simple dome, a beautiful example of which is shown in Krishna Teli's Shiva Mandir at Mairow Dih in the Chapra Subdivision of Saran District (Plate 16). But that neither the round tower nor the dome can be influenced by any principle of rythm such as we have witnessed in our earlier examples is obvious, and here we see that the side of the cella wall is treated simply as in the case of the primitive shrines at Sonpur with which we began our series as a whole.

One ugly temple at Kakraul in Darbhanga District shows an unadmirable development of this dome into a square form, but this I will not illustrate. Plate 18, however, shows a more graceful modification, where the round or square dome has passed into the octagon, this Ram Mandir at Samastipur in Saran being typical of a not numerous but picturesque class of temples, for some curious reason specially favoured and approved at Parsa in the same district, where there are several temples of this special type.

So far as my present survey goes, no further development of the simple tripartite unit is traceable in Tirhut. We have not yet by any means completed the story of temple development in Northern India, but part of the remaining story is traceable in multiplications of the units already seen. Thus (Plate 20), the Har-Mandir at Harauli in Muzaffarpur, shows us what was the next step forward, the building of two of the now familiar tripartite units side by side; in this particular case they are essentially two distinct units independent save for their being juxtaposel. But as all of you are aware, temples of this dual type frequently have the entrance porch in common, when their architectural unity is more apparent. In the Harauli temple we see the tower treated vertically by the application of two miniatures,

exactly as in the Har-Mandir at Ghatāru which was our third type above, and here too we see that there are two projections on the side of the cella wall, so that the unit even here is true to type.

Plate 22, the Shiva and Thakur Mandirs at Srinagar in Săran, show the developed tower in circular form rising from a polygonal base, where the tower itself is decorated horizontally by bands of very schematic miniatures. In essence, however, the number of these miniatures corresponds exactly to the number of surfaces in the cella wall, so that strictly speaking this temple is only another and developed form of the type preceding. The treatment is really again a vertical one, and the resulting appearance of horizontal banding is more an accident than otherwise, which is interesting as showing how two seemingly quite distinct forms of decoration can overlap.

From two such units side by side the next step is clearly to a threefold form, and Mahant Jai Ram Dass-ji's temple in Chapra (Plate 23) will illustrate this curious but pleasing stage of the development. I regret to say that this triple temple is unique in all Tirhut. Before leaving it let us appreciate the clever and successful way in which the artist has introduced variety in his treatment of the several spires.

So far as these simple units go, this is the whole series for Tirhut. But this growth into duplex and triplex forms was not the only development which took place. We have seen above that all these various temples consist of three parts, the cella, the tower and then the porch. But in some instances we see, as in the Kankali Devi temple at Simrãongarh in Nepālese territory, (Plate 25), that in course of time the familiar porch developed into a sort of verandah all around the shrine, which gives us quite a new form altogether. Here the tower is decorated vertically as in the first group above, and the same varieties of tower may be traced here as in the case of simple units. It would however serve no useful purpose to illustrate all the now familiar stages over again, and I will show you only the Shiva Mandir at Sheohar in Muzaffarpur (Plate 28), as illustrating a temple of this general class with the tower treated in the horizontal fashion

« PreviousContinue »