ia round the open mouth drawing the water to them. Some
(the Heliozoa) remain almost motionless, shooting out sticky rays of
their matter on every side to catch the food. Some form tubes to live
in; some (Coleps) develop horny plates for armour; and others develop
projectiles to pierce their prey (stinging threads).

This miniature world is full of evolutionary interest, but it is too
vast for detailed study here. We will take one group, which we know
to have been already developed in the Cambrian, and let a study of its
development stand for all. In every lecture or book on "the beauties of
the microscope" we find, and are generally greatly puzzled by, minute
shells of remarkable grace and beauty that are formed by some of these
very elementary animals They are the Radiolaria (with flinty shells, as
a rule) and the Thalamophora (with chalk frames). Evolution furnishes a
simple key to their remarkable structure.

As we saw, one of the early requirements to be fostered by natural
selection in the Archaean struggle for life was a "thick skin," and
the thick skin had to be porous to let the animal shoot out its viscid
substance in rays and earn its living. This stage above the Amoeba
is beautifully illustrated in the sun-animalcules (Heliozoa). Now the
lowest types of Radiolaria are of this character. They have no shell
or framework at all. The next stage is for the little animal to develop
fine irregular threads of flint in its skin, a much better security
against the animal-eater. These animalcules, it must be recollected,
are bits of almost pure plasm, and, as they live in crowds, dividing
and subdividing, but never dying, make excellent mouthfuls for a small
feeder. Those with the more flint in their skins were the more apt to
survive and "breed." The threads of flint increase until they form a
sort of thorn-thicket round a little social group, or a complete lattice
round an individual body. Next, spikes or spines jut out from the
lattice, partly for additional protection, partly to keep the little
body afloat at the surface of the sea. In this way we get a bewildering
variety and increasing complexity of forms, ascending in four divergent
lines from the naked ancestral type to the extreme grace and intricacy
of the Calocyclas monumentum or the Lychnaspis miranda. These, however,
are rare specimens in the 4000 species of Radiolaria. I have hundreds of
them, on microscopic slides, which have no beauty and little regular