these poisoned parts, the sympathetic nerves. For these reasons I believe that in Anolis the inward migration is a process which is ordinarily under the control of the chromatophore itself and that the outward migration, which takes place all over the animal when even only a small spot in the skin is illuminated (Parker and Starratt, 1905), is dependent upon the action of sympathetic nerves.
In the true Chameleon, as Brücke (1852) and many others have demonstrated, precisely the reverse is true; the outward migration is independent of nerves and the inward migration is produced by them. Moreover, judging from the results of experiments on the spinal cord, the nerves which in Chameleon are concerned with these changes are not sympathetic nerves, but spinal nerves.
These differences between Anolis and Chameleon I believe to be well founded. In my opinion both animals have descended from a stock in which the chromatophores were entirely independent of nervous control and in the process of descent the chromatophores of different lines became separately appropriated as effectors of the nervous system. In the ancestors of Anolis the sympathetic nervous system became related to the outward migration of pigment; and in those of the Chameleon the spinal system associated itself with the inward migration. The fact that Chameleon and Anolis belong not only to separate families, but to separate suborders of lizards, rather emphasizes this view than otherwise.
Such instances as the independent retinal chromatophores of Palæmonetes and the nervously dependent chromatophores of Chameleon and Anolis lead me to believe that chromatophores are effectors evolved independently of nervous control, but in some cases secondarily appropriated as nervous end-organs.
What has been said of chromatophores so far as their relation to nerves is concerned is probably also true of glands. The majority of glands are unquestionably independent of direct nervous control. In almost all instances a blood supply is essential to the action of a gland, and as this can be controlled by nerves there is thus an indirect influence of the nervous system on the action of the gland, but this nervous control over the blood supply is very different from a direct nervous control over secretion. I know of no good reason to assume that nerves have any direct influence on the secretions of the kidneys, the liver or even the pancreas. The pancreatic juice which appears with such precision on the arrival of food in the small intestine has been shown by Bayliss and Starling (1904) to be secreted not through the action of nerves on the gland, but through the action of a substance, secretin, produced by the food in the intestine and carried by the blood to the gland. If into the blood of a fasting animal whose nerves to the pancreas have been cut a small amount of secretin is injected, the pancreas will begin to produce its characteristic secretion.
