THE original investigator of Nature soon learns by constant experience that descriptions or even drawings, however correct, do not exactly represent the objects themselves, but are imperfect and ideal abstractions. This is true, to a greater or less extent, of every drawing of the simplest organ or tissue, and of every description of a species or genus of animals or plants; but it is especially and most emphatically true of all attempts at definitions of the larger and more comprehensive groups of organisms.
A definition of such a group as an order or class of animals, attempting as it does to state in a few words the characteristics which are common to all the forms included, is necessarily abstract, and may not, in fact cannot, be exactly embodied in any one individual of the whole group. Then, too, certain characteristics which are exhibited by only one or two aberrant forms, and are accordingly not characteristic of the group as a whole, may be omitted from the definition, although they furnish the clew to the relationship with allied groups, and are therefore of the utmost importance. An illustration which is not drawn from the organic world may make this more evident. The fact that printed books have followed and are a perfected form of the parchment manuscripts of the middle ages is shown by the ornamental initial letters, imitations of the illuminated letters of the manuscripts, which are placed at the heads of the chapters of a few books. Notwithstanding their significance, these initial letters would not find a place in any definition or general description of a modern book.
As a consequence of this inevitable lack of agreement between natural objects and their definitions, all knowledge of Nature is of very little value unless it is based upon a direct personal acquaintance with Nature itself. How different, for instance, will be that conception of such a group as the Cœlenterate, which is formed by the study of a short, definite, verbal description, from the idea in the mind of the student whose knowledge of the group as a whole has been acquired gradually by the study and comparison of the various forms of life which it includes!
Definitions are valuable and indispensable aids to study, and, as long as their necessary lack of agreement with the reality is kept in mind, they can do no harm; but the history of science warns us to be constantly on our guard, lest distinctions which seem to be sharp and absolute, when stated in words, come to be regarded as having as real an existence in Nature as in words, and we thus come, in the words of Bacon, to exchange "things for words, reason for insanity, and the world for a fable."