CHAP. II] THE MAMMALIA 33 



that it does not prove that the classification is bad or deceptive, 

 but it shews rather that animals have been evolved without regard 

 to any such conventional system of classification. 



Here we find for consideration two points of view ; of these we may take 

 first the older: this, though it needs much qualification, is far from being 

 entirely erroneous. 



A. Series continuous: intermediate forms "ancestral." On the hypo- 

 thesis that the method of the origin of species is an evolution, it follows 

 quite naturally that animal forms should merge into one another by small 

 gradations, and that classification, or grouping in classes, is actually only 

 possible in view of the fact that large numbers of animal forms have failed 

 to maintain their places in the struggle for existence. Had they not failed, 

 it would be possible to collect a demonstration-series of animal forms 

 ranging from the Amoeba to Man without any break or interruption. 

 To-day the series is discontinuous and incomplete, and the systematist seizes 

 t>n isolated groups, giving each a special name in his classification. And what 

 of the animals between these groups ? Some, having failed in the struggle for 

 existence, can only be directly known to us by such of their parts as have 

 been preserved in a fossil form. The skeletal parts only are as a rule thus 

 preserved, and incidentally this shews the importance of osteology in morpho- 

 logical study. Such are the animal "links" which have been referred to as 

 " missing." A few intermediate forms have pei-sisted down to our time, and 

 these animals prove hard to fit into a rigid system of classification. 



B. Series discontinuous : intermediate forms not necessarily ancestral. 

 Another view results from the cogency of two important objections brought 

 against that just explained (A). These objections have been stated so per- 

 fectly by Galton {Natural Inheritance, p. 32) that here only a few words will 

 be added in order to indicate their nature. In 1889, we find Galton objecting 

 to the assumption that the gradations must necessarily be small or impercep- 

 tible. He laid stress on the clear and abundant evidence "not only of the 

 appearance of considerable sports, but of their remarkable stability in here- 

 ditary transmission." And to-day, the expressions "discontinuous variation,' 1 

 and "mutation" (indicative of wide gaps between the successive forms and of 

 differences in regard to their stability in descent), are continually recurrent. 



Not only is the ground thus cut away from the claim that gradations are 

 small in every case, but a far more important sequel follows. To explain the 

 origin of small variations has seemed easier than to account for the wider 

 "sports" ; and for this some reason exists, since in the "smaller" cases, an 

 appeal can be made to comparatively familiar (if not intelligible) factors of 

 environment, such as climate or nutrition and their variations. These seem 

 inadequate to account for "sports," and until a natural basis explanatory of 

 the appearance of the latter is provided, no demonstration of "descent by 

 evolution" can be considered flawless. 



The first objection thus reminds us that evolution proceeds at a variable 

 D. M. 3 



