166 ESSAY ON CLASSIFICATION 



few natural families of animals are well characterized, or charac- 

 terized at all, we cannot open a modern treatise upon any class of 

 animals without finding the genera more or less naturally grouped 

 together, under the heading of a generic name with a termination 

 in idee or ina indicating family and sub-family distinctions; and 

 most of these groups, however unequal in absolute value, are really 

 natural groups, though far from designating always natural families, 

 being as often orders or sub-orders as families or sub-families. Yet 

 they indicate the facility there is, almost without study, to point out 

 the intermediate natural groups between the classes and the genera. 

 This arises in my opinion from the fact that family resemblance in 

 the animal kingdom is most strikingly expressed in the general form, 

 and that form is an element which falls most easily under our per- 

 ception, even when the observation is made superficially. But at the 

 same time, form is most difficult to describe accurately, and hence 

 the imperfection of most of our family characteristics and the con- 

 stant substitution for such characters of features which are not es- 

 sential to the family. To prove the correctness of this view I would 

 only appeal to the experience of every naturalist. When we see new 

 animals, does not the first glance, that is, the first impression made 

 upon us by their form, give us at once a very correct idea of their 

 nearest relationship? We perceive, before examining any structural 

 character, whether a Beetle is a Carabicine, a Longicorn, an Elaterid, 

 a Curculionid, a Chrysomeline; whether a Moth is a Noctuelite, a 

 Geometrid, a Pyralid, etc.; whether a bird is a Dove, a Swallow, a 

 Humming-bird, a Woodpecker, a Snipe, a Heron, etc., etc. But be- 

 fore we can ascertain its genus we have to study the structure of 

 some characteristic parts; before we can combine families into natural 

 groups we have to make a thorough investigation of their whole 

 structure and compare it with that of other families. So form is 

 characteristic of families; and I can add, from a careful investigation 

 of the subject for several years past, during which I have reviewed the 

 whole animal kingdom with reference to this and other topics con- 

 nected with classification, that form is the essential characteristic 

 of families. ^^ I do not mean the mere outline, but form as determined 



"These investigations, which have led to most interesting results, have delayed thus 

 far the publication of the systematic part of the Principles of Zoology (Pt. I, 1848), 

 undertaken in common with my friend, Dr. Augustus A. Gould, and which I would 

 not allow to appear before I could revise the whole animal kingdom in this new light. 



