If we were referring to insects in general, and not to these Phasmidæ in particular, we should not lay such stress on the probability of their enemies in the past being largely reptiles[1] and birds. No one who has collected insects beneath an electric light, as I have frequently done at Pretoria, attended in the same pursuit with the shadowy rushes of Bats above, and a host of patient Batrachians beneath, can doubt what wholesale insect destroyers are found in the ranks of the Chiroptera and Amphibia. But although I have found all orders of insects attracted by these lights, including Orthoptera—comprising Mantidæ, Achetidæ, Forficulidæ, Blattidæ, Gryllidæ, and Locustidæ—I personally never met with any representatives of the Phasmidæ, though of course these insects may also prove to be nocturnal in their habits, and to be also attacked by Bats. But as these animals have not been traced further back than Eocene times, we can scarcely regard them as having proved enemies to the Carboniferous Stick-insects. With the Amphibia the case is different, and, according to the late Prof. Martin Duncan, "the most ancient Amphibia appear to have first lived during the Carboniferous age, and all were tailed, had pleurodent teeth, simple in their construction.... Some were Lizard-like and others were serpentiform.... They are the Microsauria (Dawson), and the genera Hylerpeton (Owen), Hylonomus (Dawson), Brachydices (Cope), and Ophiderpeton (Huxley) are typical."[2] Here we have a host of contemporary Carboniferous enemies who may indeed have proved a great trial to the existence of unprotected Phasmidæ, and who may synchronously with the evolution of themselves have indirectly caused or induced a protective evolution in the structural form of these insects, by the mutual interdependence in those relations of cause and effect which can be expressed by the well-known appellation "natural selection." And so, for the sake of the argument, dismissing even the agency of either reptiles or birds, we still have abundant reason for believing that, though the protective resemblance of these Phasmidæ was already acquired in Carboniferous times, the presence
