. The insect returns, and
again examines the situation. Now it is a leg which is placed
cross-ways and opposes the introduction of the body; strong diseases
need strong remedies, and our _Chlorion_ sets itself to amputate this
encumbering appendage. It triumphs at last; the cockroach yields to
its efforts, and little by little penetrates the hole. As may be seen,
the labour is laborious and painful, and may present itself beneath
various aspects which call for a certain ingenuity on the part of the
animal.

[75] Reaumur, _Memoires pour servir a l'histoire des
Insectes_, Paris, 1742, t. vi., pp. 282-284.

Up to recent years the _Cerceris_ was considered to act with as much
certainty as the _Sphex_, and to obey an infallible instinct which
always guided it for the best in the interests of its offspring. The
insects it attacks belong to the genus _Buprestis_. It consumes them
in considerable numbers. Its manner of action, as described by Leon
Dufour,[76] much resembles that of the _Sphex_, and it would be
superfluous to describe it. The only fact which I wish to mention, and
which has been put out of doubt by the illustrious naturalist, is
this: the _Buprestis_ are paralysed, not dead; all the joints of the
antennae and legs remain flexible and the intestines in good condition.
He was able to dissect some which had been in a state of lethargy for
at least a week or a fortnight, although, under normal conditions,
these insects in summer decay rapidly, and after forty-eight hours
cannot be used for anatomical purposes. Another observer, Paul
Marchal, took up this question afresh, and the results which he
obtained seemed to indicate an instinct much less firm than earlier
studies tended to show.[77]

[76] "Histoire des _Cerceris_," _Ann. Sc. Nat._, ii. Serie,
t. xv., 1841, pp. 353-370.

[77] _Arch. de Zool. exp._, 1887.

_Genera less skilful in the art of paralysing victims._--These
researches show us that in the _Cerceris_ instinct is still subject to
defect. In some neighbouring genera we can seize it, as it were, in
process of formation. The way in which the _Bembex_, or Sand Wasp,
provisions burrows by maternal foresight is much less mechanical than
that of the _Sphex_. It is again Fabre who has described with most
care the customs of this hymenopterous insect.[78] It hollows out for
each egg a chamber communicating with the air by a gallery, and
performs this work with little care and very