CLASSIFICATION 27 
of the skate may be given as a possible example of a rudi- 
mentary organ. Secondly, an organ may change its function 
or, in other words, may lose its primary function but be 
preserved and greatly modified by acquiring another function. 
The skin-armour of placoid scales in sharks is not found as 
such in higher vertebrates, except the few in the neighbour- 
hood of the jaws, which form teeth. Again, the appendages 
of the crayfish show every step in modification from the 
primitive biramous swimming organ to the leg, jaw, or feeler, 
in accordance with the various functions they have acquired, 
Classification.—Hence we have seen that the animal 
kingdom forms an ascending series of organisms of struc- 
tural complexity, which is due to three kinds of gradations. 
Firstly, animals show a gradation in symmetry from the 
simple centro-symmetry to the complex plano-symmetry. 
Secondly, they show a gradation in construction from simple 
cells to many-layered individuals. Thirdly, they show a 
gradation in structure due to the functional division of 
labour. If these gradations were absolute we could form 
no classification. It would be impossible to divide the 
animal kingdom into groups if it presented a continuous 
gradation in structural characters. The breaks in structural 
sequence permit us to define certain animals and to separ- 
ate them from certain others. 
Whilst our classification is based primarily upon structural characters 
there is an important reservation. We have seen in the introduction 
that structural similarity is called omology and that there are two kinds 
of homology, inherited and ‘acquired. The acquired homology is often 
very difficult to distinguish from the inherited homology, but the ideal 
classification to which all zoologists aspire is based purely upon inherited 
homology or upon homogenetic characters ; if we place together in one 
group a number of individuals because they have omogenous similarity 
in structure, we shall by our definition be correlating animals which 
are descended from a common stock. This is a satura classification, 
for in it we strive to give expression to the natural relationships of the 
animals. Let us take a very simple example. If we decide to put 
in one group the animals which swim in the sea, have a tail-fin and 
pectoral fins and are of a fish-like shape, we create a group containing 
the whales and fishes. This is an artdficda/ classification, for further 
examination shows that the whale agrees with land-mammals in nearly 
all the most important mammalian characters and that its fish-like shape 
is acgutred or due to adaptation to an aquatic life. 
The determination of natural affinities is largely helped by the study 
of embryology and of paleontology, but there is no exact criterion for 
