field naturalists on these points tends to prove how partial or moderate must be the danger in the present day, and how considerably more intense it must have been in some former time to have prompted the evolution of the wonderfully simulating guises, which we can only conceive as evolved for protective purposes.
A repetition of observations will frequently qualify the premises on which many conclusions are based. Many recorded facts are of course utterly erroneous. Thus in 1666 Schefferus records in the 'Philosophical Transactions' that Swallows sink into lakes in autumn, and hibernate in a manner precisely similar to Frogs. In 1741 Fermier-Général Witkowski made legal testimony to the effect that two Swallows had been taken from a pond at Didlacken in his presence in a torpid state; that they eventually regained animation, and after fluttering about, died some three hours after their capture. In 1748 the great Swedish chemist Wallerius wrote that he had on several occasions seen Swallows clustering on a reed until they all disappeared beneath the surface.[1] Thus a traveller in a tropical forest might from paucity of observation form a wrong impression as to the relation of the liane and the stem or tree to which it is attached. He would frequently find "the hard basal parts of a liane stem twisted and coiled apparently around nothing. This is due to the fact that the original support had been killed, and then, slowly rotting into dust, has been denuded away by the wind and rain." Our traveller might then record the murderous action of lianes as of a somewhat universal character. But further observations would show the action quite reversed. As Kerner describes the process: "If the erect young stem is stronger and more vigorous than the twiner which encircles it, which has been used as a prop, it does not allow itself to be strangled; the twiner is destroyed when they both increase in thickness. The coils of the climber are gradually stretched tighter and tighter, and many are the contrivances which exist for preventing the tension from immediately acting injuriously on the movement of the sap in the interior of the twining liane stem. As this thickening continues, the pull on the coils becomes so great that the death of the liane results."[2]
