s the
cause of their development or arrest of development. Lastly, I will
treat of the natural order of animals, and show what should be their
fittest classification and arrangement."[212]

It seems unnecessary to give Lamarck's intentions with regard to his
second and third parts, as they do not here concern us; they deal with
the origin of life and mind.

The first chapter of the work opens with the importance of bearing in
mind the difference between the conventional and the natural, that is to
say, between words and things. Here, as indeed largely throughout this
part of his work, he follows Buffon, by whom he is evidently influenced.

"The conventional deals with systems of arrangement, classification,
orders, families, genera, and the nomenclature, whether of different
sections or of individual objects.

"An arrangement should be called systematic, or arbitrary, when it does
not conform to the genealogical order taken by nature in the development
of the things arranged, and when, by consequence it is not founded upon
well-considered analogies. There is such a thing as a natural order in
every department of nature; it is the order in which its several
component items have been successively developed.[213]

"Some lines certainly seem to have been drawn by Nature herself. It was
hard to believe that mammals, for example, and birds, were not
well-defined classes. Nevertheless the sharpness of definition was an
illusion, and due only to our limited knowledge. The ornithorhynchus and
the echidna bridge the gulf.[214]

"Simplicity is the main end of any classification. If all the races, or
as they are called, species, of any kingdom were perfectly known, and if
the true analogies between each species, and between the groups which
species form, were also known, so that their approximations to each
other and the position of the several groups were in conformity with the
natural analogies between them--then classes, orders, sections, and
genera would be families, larger or smaller; for each division would be
a greater or smaller section of a natural order or sequence.[215] But in
this case it would be very difficult to assign the limits of each
division; they would be continually subjected to arbitrary alteration,
and agreement would only exist where plain and palpable gaps were
manifest in our series. Happily, however, for classifiers there are, and
will always probably remain, a number of unknown forms."[216]

Tha