When we examine closely these different signs or characters so deeply engraved [pecked] on the rock, doubtless by means of some hard stone (obsidian or basalt), the first thing observed is that several identical signs are reproduced several times in the same group. These are, first, round and oval characters, more or less perfect, sometimes simple and isolated, again agglomerated in one group. These characters so often reproduced are again seen in juxtaposition or united, sometimes to others which are similar, sometimes to different ones, and even inclosed in others similar to them; for example, _a_ in Fig. 144. Round or more or less oval characters reappear several times in _b_. Others, which are not met with more than once or twice among the groups of signs, also present notable variations; examples in _c_. Of these are formed composite groups _d_, which belong, however, to the system of round signs. Other analogous but not identical signs appear to assume rather the ovoid form than the round, and seem to have been so traced as not to be confounded with the round symbols. Some of them resemble leaves or fruit. Another system of simple characters is the straight line, which can be represented by a stroke of the pen, isolated or repeated as if in numeration, and sometimes accompanied by other signs. Other peculiar signs shown in _e_, which are not repeated, figure in the different groups of characters which the author has reproduced. We notice further, in _f_, a small number of signs which bear a certain analogy to each other, and several of which are accompanied by other and more simple characters. Several others still more complicated are in eccentric shapes which it is attempted to present in _g_. Including the common oval characters often repeated and those consisting of a simple stroke similar to the strokes made by school children, all the various engraved characters scarcely exceed 400. Fig. 145 gives a view of a series of different groups of signs in the length of the whole lava flow. The copyist has expressed by dots those symbols which were confused, partly defaced by the weather, or destroyed by fissures in the rock. [Illustration: FIG. 145.--Petroglyphs in Canary Islands.] The same author (_b_) gives an account of several strange characters found engraved on a rock of the grotto of Belmaco, in the island of La Palma, one of the Canaries. He says: