these in colour, slender
shape of body and limbs, rapid movements, and the readiness with which
they take to flight. On Dec. 21, 1812, Burchell captured one such beetle
(Promeces viridis) at Kosi Fountain on the journey from the source
of the Kuruman River to Klaarwater. It is correctly placed among the
Longicorns in his catalogue, but opposite to its number is the comment
"Sphex! totus purpureus."

In our own country the black-and-yellow colouring of many stinging
insects, especially the ordinary wasps, affords perhaps the commonest
model for mimicry. It is reproduced with more or less accuracy on moths,
flies and beetles. Among the latter it is again a Longicorn which offers
one of the best-known, although by no means one of the most perfect,
examples. The appearance of the well-known "wasp-beetle" (Clytus
arietis) in the living state is sufficiently suggestive to prevent
the great majority of people from touching it. In Burchell's Brazilian
collection there is a nearly allied species (Neoclytus curvatus) which
appears to be somewhat less wasp-like than the British beetle. The
specimen bears the number "1188," and the date March 27, 1827, when
Burchell was collecting in the neighbourhood of San Paulo. Turning to
the corresponding number in the Brazilian note-book we find this
record: "It runs rapidly like an ichneumon or wasp, of which it has the
appearance."

The formidable, well-defended ants are as freely mimicked by other
insects as the sand-wasps, ordinary wasps and bees. Thus on February
17, 1901, Guy A.K. Marshall captured, near Salisbury, Mashonaland,
three similar species of ants (Hymenoptera) with a bug (Hemiptera) and
a Locustid (Orthoptera), the two latter mimicking the former. All the
insects, seven in number, were caught on a single plant, a small bushy
vetch. ("Trans. Ent. Soc. Lond." 1902, page 535, plate XIX. figs.
53-59.)

This is an interesting recent example from South Africa, and large
numbers of others might be added--the observations of many naturalists
in many lands; but nearly all of them known since that general awakening
of interest in the subject which was inspired by the great hypotheses
of H.W. Bates and Fritz Muller. We find, however, that Burchell had
more than once recorded the mimetic resemblance to ants. An extremely
ant-like bug (the larva of a species of Alydus) in his Brazilian
collection is labelled "1141," with the date December 8, 1826, when
Burchell was at the Rio das Pedr