, for lack of the requisite
flexibility. This total absence of movement in the tertiary larva,
when a few days old, and also in the nymph, together with the
smallness of the space left free in the shell, would necessarily lead
to the conviction, if we had not witnessed the first moments of the
tertiary larva, that it is absolutely impossible for the creature to
turn right round.

And now see to what curious inferences this lack of observations made
at the due moment may lead us. We collect some pseudochrysalids and
heap them in a glass jar in all possible positions. The favourable
season arrives; and with very legitimate astonishment we find that, in
a large number of shells, the larva or nymph occupies an inverted
position, that is to say, the head is turned towards the anal
extremity of the shell. In vain we watch these reversed bodies for any
indications of movement; in vain we place the shells in every
imaginable position, to see if the creature will turn round; in vain,
once more, we ask ourselves where the free space is which this turning
would demand. The illusion is complete: I have been taken in by it
myself; and for two years I indulged in the wildest conjectures to
account for this lack of correspondence between the shell and its
contents, to explain, in short, a fact which is inexplicable once the
propitious moment has passed.

On the natural site, in the cells of the Anthophora, this apparent
anomaly never occurs, because the secondary larva, when on the point
of transformation into the pseudochrysalis, is always careful to place
its head uppermost, according as the axis of the cell more or less
nearly approaches the vertical. But, when the pseudochrysalids are
placed higgledy-piggledy in a box or jar, all those which are upside
down will later contain inverted larvae or nymphs.

After four changes of form so profound as those which I have
described, one might reasonably expect to find some modifications of
the internal organization. Nevertheless, nothing is changed; the
nervous system is the same in the tertiary larva as in the earlier
phases; the reproductive organs do not yet show; and there is no need
to mention the digestive apparatus, which remains invariable even in
the perfect insect.

The duration of the tertiary larva is a bare four or five weeks, which
is also about the duration of the second. In July, when the secondary
larva passes into the pseudochrysalid stage, the tertiary larva passes