Page:The Zoologist, 4th series, vol 3 (1899).djvu/492

462 that is because they mistake it for a Hawk, for the longer I observe the more I am convinced that birds and animals often act from causes quite distinct from those which at first sight appears sufficient to account for their motions." The dread experienced by small birds for their larger brethren of prey is probably open to qualification, for Gilbert White tells us of a Swallow who "built its nest on the wings and body of an Owl that happened by accident to hang dead and dry from the rafters of a barn."

As with "mimicry," so the theory of "warning colours" may be hastily predicated. Among Flatworms in the terricolous Triclads or Land Planarians, some species "are frequently banded or striped with brilliant colours. Geoplana cærulea, Mos., has a blue ventral surface, and is olive-green or dark Prussian blue above. G. splendens, Dendy, is marked dorsally by three stripes of emerald-green alternating with four dark brown longitudinal bands. The mode of colouration, though somewhat variable, is an important specific character. Its significance, however, is not clearly understood. The colours may be a warning signal, as some Geoplana at least are disagreeable to the taste of man and some birds; but since Land Planarians are largely nocturnal animals, living by day under logs, banana-leaves, and in other moist and dark situations, this explanation is clearly insufficient." Among the Polychaete Worms the same caution is necessary. "Carnivorous forms like Amphinomids and Syllids present as wide a range of tint as the limnivorous forms, like Cirratulus, Sabella, or Maldanids. Shore-lovers and deep-sea dwellers and surface-swimmers all exhibit equally bright or equally sombre tints; it is therefore difficult and rash to dogmatise on the 'use' of these colourings to these animals, or to point to this worm as being protectively, to the other as being warningly, coloured; for we are too ignorant as to the habits of the worms."

As we record instances of what appear only capable of being ascribed to "suggestive but mistaken mimicry," we meet with natural resemblances which seem to fall under a category of