The sudden disappearance of groups of animals, which have existed through long periods of time and have attained a high degree of specialization, is a phenomenon of which many instances will occur to every student of Palæozoology. For instance, to mention only two cases in illustration, we may refer to the disappearance of the Dinosaurs at the end of the Secondary period, and that of the North American Titanotheres in the Miocene. Of the proximate causes of this extinction little is known: they must have been either inherent in the organisms themselves, or have been connected with the relations of the organisms with their environment; probably in every case several factors co-operated to bring about the observed result. In a recent paper by Mr. C. B. Crampton (Proc. Roy. Phys. Soc. Edinburgh, vol. xiv, p. 461) a possible inherent cause of extinction is suggested. It is impossible to do justice to this interesting paper in a short note, but the gist of the argument seems to be as follows:—In the original unicellular organism the possibilities of variation are almost infinite, but as soon as evolution along any line begins, these possibilities are restricted, and become more and more so the more highly specialized the animal is; in short, the potential variation of an organism becomes less and less as specialization advances. Furthermore, under the influence of natural selection, in each generation the individuals which tend to vary in the same direction will survive, while at the same time, as already pointed out, their capacity for variation becomes more and more restricted.