Some authors as Lamarck, Whewell &c., believe that the degree of affinity on the natural system depends on the degrees of resemblance in organs more or less physiologically important for the preservation of life. This scale of importance in the organs is admitted to be of difficult discovery. But quite independent of this, the proposition, as a general rule, must be rejected as false; though it may be partially true. For it is universally admitted that the same part or organ, which is of the highest service in classification in one group, is of very little use in another group, though in both groups, as far as we can see, the part or organ is of equal physiological importance: moreover, characters quite unimportant physiologically, such as whether the covering of the body consists of hair or feathers, whether the nostrils communicated with the mouth{432} &c., &c., are of the highest generality in classification; even colour, which is so inconstant in many species, will sometimes well characterise even a whole group of species. Lastly, the fact, that no one character is of so much importance in determining to what great group an organism belongs, as the forms through which the embryo{433} passes from the germ upwards to maturity, cannot be reconciled with the idea that natural classification follows according to the degrees of resemblance in the parts of most physiological importance. The affinity of the common rock-barnacle with the Crustaceans can hardly be perceived in more than a single character in its mature state, but whilst young, locomotive, and furnished with eyes, its affinity cannot be mistaken{434}. The cause of the greater value of characters, drawn from the early stages of life, can, as we shall in a succeeding chapter see, be in a considerable degree explained, on the theory of descent, although inexplicable on the views of the creationist. {432} These instances occur with others in the _Origin_, Ed. i. p. 416, vi. p. 572. {433} _Origin_, Ed. i. p. 418, vi. p. 574. {434} _Origin_, Ed. i. pp. 419, 440, vi. pp. 575, 606.