age may puzzle men who happen to ponder seriously on first causes. I
recall an afternoon when such playthings were being manufactured
abundantly. Globular, oval, and sausage-shaped dollops of dark-grey mud
were twirling and rolling on the fringe of listless wavelets. The
uniformity of the several models and their apparent solidity excited
curiosity. Upon investigation all the large examples were more or less
coated with sand. Some were so completely and smoothly enveloped that
they appeared to be actual balls of sand and shell grit. The mass,
however, was found to be mud mixed with fine sand, with generally a
shell or portion thereof, or a fragment of coral as a kernel or core. In
fact, each of the dollops was a fair sample of the material of the ocean
floor extending from the inner edge of the coral to the beach.

With so many samples in view one could observe the whole process of
formation. The crescentic sweep of the wavelets rolled fragments of shell
or coral in the mud, successive revolutions adding to the respective
bulks by accretion. As the tide rose each piece was trundled on to the
sloping beach, to be rolled and compressed until coated with a mosaic of
white shell chips, angularities of silica and micaceous spangles, the
finished article being cast aside as the tide receded.

Sometimes the wavelets did the kneading and rolling so clumsily that the
nodule was malformed, but the majority were singularly symmetrical,
evidencing nice adjustment between the degree of adhesiveness of the
"pug" and the applied force of the wave. Several weighed nearly a quarter
of a pound, while the majority were not much bigger than marbles, and the
oval was the most frequent form.

Is it reasonable to conjecture that some of these singular formations
which Neptune turned out by the score during an idle afternoon may be