[21] See _Proc. Zool. Soc._ June 4, 1889, for an account of the performances in this respect of the Chimpanzee "Sally." Also, for some remarks on the psychology of the subject, in _Mental Evolution in Man_, p. 215. I should like to take this opportunity of stating that, after the two publications above referred to, this animal's instruction was continued, and that, before her death, her "counting" extended as far as ten. That is to say, any number of straws asked for from one to ten would always be correctly given. Once more, it is obviously unfair to select such men as a "Newton, a La Place, a Gauss, or a Cayley" for the purpose of estimating the difference between savages and civilized man in regard to the latter "faculty." These men are the picked mathematicians of centuries. Therefore they are men who not only enjoyed all the highest possible benefits of individual culture, but likewise those who have been most endowed with mathematical power congenitally. So to speak, they are the best variations in this particular direction which our race is known to have produced. But had such variations arisen among savages it is sufficiently obvious that they could have come to nothing. Therefore, it is the _normal average_ of "mathematical faculty" in civilized man that should be contrasted with that of savage man; and, when due regard is paid to the all-important consideration which immediately follows, I cannot feel that the contrast presents any difficulty to the theory of human evolution by natural causation.