Pieris surpasses us. She does not consult the
seed-vessel, to see if it be long or short, nor yet the petals, four in
number and arranged in a cross, because the plant, as a rule, is not in
flower; and still she recognizes offhand what suits her caterpillars,
in spite of profound differences that would embarrass any but a
botanical expert.

Unless the Pieris has an innate power of discrimination to guide her,
it is impossible to understand the great extent of her vegetable realm.
She needs for her family Cruciferae, nothing but Cruciferae; and she
knows this group of plants to perfection. I have been an enthusiastic
botanist for half a century and more. Nevertheless, to discover if this
or that plant, new to me, is or is not one of the Cruciferae, in the
absence of flowers and fruits I should have more faith in the
Butterfly's statements than in all the learned records of the books.
Where science is apt to make mistakes instinct is infallible.

The Pieris has two families a year: one in April and May, the other in
September. The cabbage-patches are renewed in those same months. The
Butterfly's calendar tallies with the gardener's: the moment that
provisions are in sight, consumers are forthcoming for the feast.

The eggs are a bright orange-yellow and do not lack prettiness when
examined under the lens. They are blunted cones, ranged side by side on
their round base and adorned with finely-scored longitudinal ridges.
They are collected in slabs, sometimes on the upper surface, when the
leaf that serves as a support is spread wide, sometimes on the lower
surface when the leaf is pressed to the next ones. Their number varies
considerably. Slabs of a couple of hundred are pretty frequent;
isolated eggs, or eggs collected in small groups, are, on the contrary,
rare. The mother's output is affected by the degree of quietness at the
moment of laying.

The outer circumference of the group is irregularly formed, but the
inside presents a certain order. The eggs are here arranged in straight
rows backing against one another in such a way that each egg finds a
double support in the preceding row. This alternation, without being of
an irreproachable precision, gives a fairly stable equilibrium to the
whole.

To see the mother at her laying is no easy matter: when examined too
closely, the Pieris decamps at once. The structure of the work,
however, reveals the order of the operations pretty clearly. The
ovipositor swings slow