These avenues are of two classes. First, those leading to circles. About the origin of this class there can be very little hesitation. They represent externally the passages in tumuli which lead to the central chamber; take, for instance, this example from a now destroyed[66] tumulus near St. Helier, in Jersey.[67] The circular chamber was 24 feet in diameter, and contained originally seven little cells, each roofed by a single slab of stone. This circular area was approached by an avenue, 17 feet long at the time of its destruction, which was roofed throughout the whole length with slabs of stone. The central chamber never, however, appears to have been vaulted, so that access to the tombs through this passage could never have been possible after the mound was finished. The chamber was found filled with earth, and the whole monument covered up by a tumulus of considerable extent. It need hardly be observed that it is more unlikely that any people should cover up such a monument at any subsequent age, than that they should dig out such monuments and leave them standing without their envelopes, as is so generally assumed. The tumulus was removed, because the officer in command of the neighbouring fort wanted a level parade-ground. As it stood uncovered it was a miniature Avebury, and the position of its cells may give us a hint where the bodies may be found there--near the outer circle of stones, where they have not been looked for. But of this hereafter. It is meanwhile evident that while these monuments were in course of erection they stood as shown in the last woodcut, and it is also tolerably clear that when people became familiar with their aspect in this state, they may have learned to regret hiding under a heap of earth what we certainly would have thought more interesting as it was. In like manner, as John Stuart well remarks, "If the cairns at New Grange were removed, the pillars would form another Callernish."[68] It is true, however, that if the Jersey monument is the type of Avebury, the latter must be comparatively modern, as a coin of Claudius, found in one of the cells at St. Helier,[69] probably fixes its date. Again, as we expect to be able to prove that New Grange is subsequent to the Christian era, Callernish must be more modern also. Be this as it may, I think there can be very little doubt that these exposed circles, with their avenues, took their rise, as in the case of dolmens, from people becoming familiar with their forms before they were covered up, and eventually reconciling themselves to dispense with the envelope. In the case of the circles, the new plan was capable of infinitely greater extension than in that of the dolmens; but the process seems to have been the same in both instances.