244 Mr. J. A. Ryder on the 



starting from any scale in any oblique row, tliey may be 

 counted either forward or backward longitudinally or in con- 

 formity with the direction of the axis of the body of the fish. 

 This is conspicuously the case in Clupeoids and some 

 Cyprinoids. 



In such archaic types as these, approximating the primitive 

 isospondylous condition, it is also found that the number of 

 scales in a longitudinal row corresponds on the sides of the 

 body very exactly with the number of muscle-plates or somites 

 of the body. It is also found that the myocommata or sheets 

 of connective tissue intervening between the successive somites 

 are attached with great firmness to the deeper layers of the 

 skin or corium. Such a construction, together with the 

 peculiar arrangement of the muscle-plates at the time the 

 scales begin to develop, conditions the further growth of the 

 scale matrix. This is affected in such a manner that the 

 whole of the integument is thrown into definitely circum- 

 scribed areolae during the ordinary movements of the fish in 

 swimming. The central portions of each of these areolge are 

 left in a quiescent condition, while their margins are wrinkled 

 or folded as a result of the current action of the lateral muscles 

 of the body. In this wise each and every one of the dermal 

 and epidermal areolae are circumscribed by the action of the 

 fish in the normal act of swimming. In each of the circum- 

 scribed areolee a scale develops ; the continuity of its develop- 

 ment with its fellows across the margins of the areola is 

 prevented by the continual bendings or flexures to which the 

 dermis is there subjected owing to the action of the muscles. 



This will be better understood by referring to the accom- 

 panying diagram, representing the arrangementof the muscular 

 somites of a Cyprinoid {Carassius) with their intervening 

 myocommata as seen from the side when the skin with its 

 scales is removed. Before proceeding further, however, it 

 may be well to insist upon the fact that the rows of scales are 

 found to conform to the successive somites. This is of itself 

 significant. The careful interpretation of the facts from 

 observation, however, discloses a very remarkable effect due 

 to the peculiar arrangement of the muscle-plates. 



As is well known, the muscular masses of the sides of the 

 body of a fish are arranged in the form of two longitudinal 

 trihedral columns separated along the middle line of the side 

 a to h into a dorsal and ventral half. The somites entering 

 into the composition of these ventral and dorsal masses were 

 at first absolutely continuous across the longitudinal horizontal 

 septum a, h. If we suppose the somites of the adult as deve- 

 loped from a continuous embryonic segment extending the 



