The greatest marvel is that such polychrome masterpieces as the bisons standing and couchant or with limbs huddled together were executed on the ceilings of inner vaults and galleries, where the light of day never penetrated. Nowhere does smoke blur their outlines, probably (as Parkyn[12] suggests) because of long oxidisation. The art of artificial illumination had evidently progressed far. We now, indeed, know that stone lamps, decorated in one case with the head of an ibex, already existed. “_Les extremes se touchent_” was here aptly exemplified, for to a very young child was it reserved to discover the very oldest art gallery in the world. In 1879 Señor de Santuola chanced to be digging in a cave on his property, when he heard his little daughter cry, “Toros, toros!” Realising quickly that this was no warning of an impending charge by bulls, he followed her gaze to the vaulted ceiling, where his eyes there espied “the finest expression of Palæolithic art extant.” This little Spanish girl was the first for many, many thousands of years to behold a collection of pictures, which demonstrate not only the high point of excellence to which the art of the Troglodytes had attained, but also, from the absence of perspective and of decorative as compared with pictorial composition, indicate how long is probably the interval and how far is the separation between them and the Men of the Neolithic Age. Not only in the character of their Art, which if more specialised in subjects was superior in representative quality, but also in the substance and in the method of fashioning their fishing and hunting implements, the separation between the Old Stone and the New Stone Man is very marked. The former for their stone implements almost always used flint. They worked it to shape merely by flaking or chipping. The latter employed also diorite, quartzite, etc., and in addition to flaking fashioned them by grinding and polishing.[13] It must, I fear, be acknowledged that the _caches_ of the New Stone Age fail to give us the help expected towards settling what was the first implement employed. It is true that they yield hooks, nets, net-sinkers, which may have been merely developments of Troglodyte tackle, but, judging from the absence of any surviving Palæolithic example, were more probably new inventions. But neither these nor the implements of succeeding Ages furnish us with evidence sufficient to decide the tackle first employed by the earliest fisherman, or even by the Old Stone Man, for, as Cartailhac truly warns us, “Ce n’est pas, comme on l’a dit à tort, le début de l’art que nous découvrons. L’art de l’âge du renne est beaucoup trop ancien.”[14]