structural characters probably of the original wild horse, have appeared. We also learn from such instances that, while races among domesticated animals may appear suddenly, they revert to the original type, when unmixed, comparatively slowly; and this especially when the variation is in the nature of degeneracy. 4. Some characters are more subject to variation than others. In the higher animals variation takes place very readily in the color and texture of the skin and its appendages. This, from its direct relation to the external world, and ready sympathy with the condition of the digestive organs, might be expected to take the lead. In those domesticated animals which are little liable to vary in other respects, as the cat and duck, the color very readily changes. Next may be placed the stature and external proportions, and the form of such appendages as the external ear and tail. All these characters are very variable in domestic animals. Next we may place the form of the skull, which, though little variable in the wild state, is nearly always changed by domestication. Psychological functions, as the so-called instincts of animals, are also very liable to change, and to have these changes perpetuated in races. Very remarkable instances of this have been collected by Sir C. Lyell[165] and Dr. Prichard. Lastly, important physiological characters, as the period of gestation, etc., and the structure of the internal organs connected with the functions of nutrition, respiration, etc., are little liable to change, and remain unaffected by the most extreme variations in other points; and it is, no doubt, in these more essential and internal parts that the tendency survives to return under favorable circumstances to the original type.