It has also been shown that certain forms which were once supposed to be especially transitional and intermediate (as, _e.g._, the aye-aye) are really by no means so; while the general rule, that the progress of forms has been "from the more general to the more special," has been shown to present remarkable exceptions, as, _e.g._, Macrauchenia, the Glyptodon, and the sabre-toothed tiger (Machairodus). Next, as to specific stability, it has been seen that there may be a {223} certain limit to normal variability, and that if changes take place they may be expected _a priori_ to be marked and considerable ones, from the facts of the inorganic world, and perhaps also of the lowest forms of the organic world. It has also been seen that with regard to minute spontaneous variations in races, there is a rapidly increasing difficulty in intensifying them, in any one direction, by ever such careful breeding. Moreover, it has appeared that different species show a tendency to variability in special directions, and probably in different degrees, and that at any rate Mr. Darwin himself concedes the existence of an internal barrier to change when he credits the goose with "a singularly inflexible organization;" also, that he admits the presence of an _internal_ proclivity to change when he speaks of "a whole organization seeming to have become plastic, and tending to depart from the parental type." We have seen also that a marked tendency to reversion does exist, inasmuch as it sometimes takes place in a striking manner, as exemplified in the white silk fowl in England, _in spite of_ careful selection in breeding. Again, we have seen that a tendency exists in nature to eliminate hybrid races, by whatever means that elimination is effected, while no similar tendency bars the way to an indefinite blending of varieties. This has also been enforced by statements as to the prepotency of certain pollen of identical species, but of distinct races. To all the preceding considerations have been added others derived from the relations of species to past time. It has been contended that we have as yet no evidence of minutely intermediate forms connecting uninterruptedly together undoubtedly distinct species. That while even "horse ancestry" fails to supply such a desideratum, in very strongly marked and exceptional kinds (such as the Ichthyosauria, Chelonia, and Anoura), the absence of links is very important and significant. For if every species, without{224} exception, has arisen by minute modifications, it seems incredible that a small percentage of such transitional forms should not have been preserved. This, of course, is especially the case as regards the marine Ichthyosauria and Plesiosauria, of which such numbers of remains have been discovered.