20. In the Vertebrates the coalescence of ganglia is such that the spinal axis is one great centre. We do indeed anatomically and physiologically subdivide it into several centres, because several portions directly innervate separate organs; but its importance lies in the intimate blending of all parts, so that _fluctuating combinations_ of its elements may arise, and varied movements result. Each centre combines various muscles; the axis is a combination of centres. The brainless frog, for instance, has still the spinal cord, and therefore the power not only of moving either of his limbs, but also of combining their separate movements: if grasped, he struggles and escapes; if pricked, he hops away. But these actions, although complex, are much less complex and varied than the actions of the normal frog. There is not only a coalescence of ganglia, but a greater and greater concentration of the substance in the upper portions of the axis. In the inferior vertebrates, and in the mammalian embryo, the spinal cord occupies the whole length of the vertebral canal from the head to the tip of the tail; and here the centres of reflexion correspond with the several segments. But as the cranial mass develops there is a withdrawal of neural substance from the lower parts, and the centres of reflexion are then some way removed from the segments they innervate. In the animal development there is even a greater and greater predominance of the upper portions, so that the brain and medulla oblongata are of infinitely more importance than the spinal cord. 21. Besides the central group of elements which belong to fixed and definite actions, we must conceive these elements capable of variable combinations, like the pieces of colored glass in a kaleidoscope, which fall into new groups, each group having its definite though temporary form. The elements constitute really a continuous network of variable forms. It is to such combinations, and not to fixed circumscribed ganglia, that we must refer the subordinate centres of the axis. We speak of a centre for Respiration, a centre for Laughing, a centre for Crying, a centre for Coughing, and so on, with as much propriety as we speak of a centre for Swallowing or for Walking. Not that in these cases there is a circumscribed mass of central substance set apart for the innervation of the several muscles employed in these actions, and for no other purpose. Each action demands a definite group of neural elements, as each geometric form in the kaleidoscope demands a definite group of pieces of glass; but these same pieces of glass will readily enter into other combinations; and in like manner the muscles active in Respiration are also active in Laughing, Coughing, etc., though differently innervated and co-ordinated.