This leads us to the subject of changeability of colours in the same individual. That such changes do occur is well known, and it is still more remarkable that they are produced in rapid succession, apparently at the will of the fish concerned; for its tints will vary as it moves from place to place so as to always harmonise with the surroundings, and also in response to other conditions. The mechanism by which such variations are produced has also been studied and explained:--The colouring matter is held in little vesicles beneath the skin, and these vesicles are capable of being compressed by muscles quite under the control of the fish. When they are globular in form the contained pigment appears dark, but when they are flattened by muscular compression, the pigment is spread over a much larger area, and thus greatly reduced in depth of tint. As with all vertebrates, the central axis of the internal skeleton of a fish consists of the backbone and the skull. The structure of the latter is so complicated, and its description so full of technicalities, that we deem it advisable to pass it over in a work like this where the scope is so large in proportion to the space available; and this we do with reluctance, because the detailed study of the skull is of real importance to those who would thoroughly understand the principles of classification. The backbone consists of a variable number of cylindrical vertebræ, united end to end to form a continuous column, both the anterior and posterior faces of each being concave. On the dorsal surface of each vertebra there is a V-shaped arch, surmounted by a spine, the former serving to protect the spinal cord, and the latter giving attachment to the muscles of the back. Some of the vertebræ are also provided with processes for the attachment of the ribs, and those of the tail possess an arch and a spine on the ventral as well as on the dorsal side. It has already been shown that the pectoral fins are jointed to a girdle. This girdle corresponds with the shoulder-blade of higher animals, and gives direct attachment to the rays of the fin, which may be regarded as the equivalent of the fingers, and thus there is no part of the limb corresponding with the arm. The pelvic fins also are frequently jointed to a pelvic girdle or hip, but this is a very rudimental structure, or is even entirely absent in some species.