3. The species which resemble or “mimic” these dominant groups, are comparatively less abundant in individuals, and are often very rare.
The second and third of these laws are confirmed by what we know of mimetic resemblances among spiders. They mimic ants much oftener than other creatures, and ants are very abundant, are specially protected, and are much more numerous than the mimetic spiders. To the first law, also, they conform to a great extent, since everything tends to show that in tropical America and in Africa the ant and the spider, the one mimicked and the other mimicking, are always found together. So far as I can discover, however, the ant-like spiders of North America are not found in company with any species of ant which they resemble. This may be because they do not mimic any particular species, but only the general ant-like form; or, considering that the genera which contain their nearest relatives are much more abundant in Central and South America, it may be that these forms were originally tropical, mimicking some tropical species of ants, and that after the Glacial Epoch they migrated northward, leaving the ants behind them. However this may be, their peculiar form has served them well, since they have maintained themselves as fairly abundant species with a lower fecundity than is found in any other group of spiders.
The cases in which one species mimics another may be divided, according to the kind of benefit derived, into four classes: Class 1. As a rule, where we find one species mimicking another, the mimicked species possesses some special means of defence against the enemies of both. This defence may consist of a disagreeable taste or odor, as in the Heliconidæ, which are mimicked by other butterflies; of some special weapon of offence, as where wasps and bees are mimicked by flies and moths, or poisonous vipers by harmless caterpillars; or of a hard shell, as where the coriaceous beetles are mimicked by those that are soft-bodied.
Instances of this rule are exceedingly numerous; indeed, Wallace says that specially protected forms are always mimicked; still we have nothing mimicking our Gasteracanthidæ.
Class 2. The mimetic may prey upon the mimicked species, its disguise enabling it to gain a near approach to its victims; as the mantis, mentioned by Bates as exactly resembling the white ants upon which it feeds; and the flies which mimic bees, upon which they are parasitic, and are thus able to enter the nests of the bees and lay eggs on the larvæ.
Class 3. The mimetic species may, by its imitation, be protected from the attacks of the creature it mimics, as is the case with the crickets and grasshoppers which mimic their deadly foe, the hunter wasp.
Class 4. The mimetic species may prey upon some creature which is found commonly with, and is not eaten by, the mimicked species.
No two of these classes are mutually destructive so that in any case of mimicry a double advantage may be gained.
Let us see which of these advantages has directed the development of mimetic tendencies among spiders.
While among beetles and butterflies we most commonly find mimicry of one species by another within the same order, we have no instance of a spider mimicking another spider. This may be accounted for by the fact that the specially protected spiders depend for their safety upon the possession of hard plates and spinous processes, and although the hardened epidermis might be imitated (we know that hard-shelled beetles are mimicked by others that are soft), spines could scarcely be imitated by a soft-bodied creature with sufficient accuracy to insure disguise.