OF THE AMAZON VALLKY. 513 



(55' "\y. 1(^115.) tho ros('inl)laiic(> is ])rrl'c'ct (I'l. J^V. i'lL;'. 1) ; :iiul this is tlic only lorm of 

 tlie Lcpldlis kuowu in tlic locality. The varieties lij^ured J'l. L\l. iius. 1, 2, ;>, show 

 different degrees of resemblance to Uhom'm Clirysodoitid (fii;'. 3 o) ; these, therefore, 

 exhibit the selection in process. Thus, althoug'li we are unable to watch the ])rocess of 

 formation of a new race as it occurs in time, we can see it, as it were, at one glance, by 

 tracing the changes a species is simultaneously undergoing in dillerent parts of the ai'ca 

 of its distribution. 



The fact of oyeof the forms of Lep/c/is Theonoi;, namely Jj. Liisinov, mimicking at Ega, 

 not an Illiomia, but a llourishing species of another (piite distinct family {Slalaclitis 

 Diivalii), shows that the object of the mimetic tendencies (jf the species is simply dis- 

 guise, and that, the simple individual diU'erences in that kjcality being originally in the 

 direction, not of an Itliomia, but of another object eciually wcdl answering the purpose, 

 selection operated in the direction of that other object. This point is well illustrated 

 by the species of a small group of Longicorn Beetles already cited, some of which mimic 

 a piece of bark, and others in.sects of another family— and l)y hunting Spidi-rs, many of 

 which wear the form of insects, and many that of inanimate objects amongst which 

 they seek their prey. 



When the persecution of a varialde local form of our LepUtUs is close or long con- 

 tinu(xl, the indeterminate variations naturally become extinct; nothing then remains in 

 that locality l)ut the one exact counterfeit, whose exactness, it must lie added, is hence- 

 forward kept up to the mark by the insect pairing necessarily with its exact counterpart, 

 or breeding /;/ (nid in. This is the condition of LcptdHs Theonoi' (i'l. LV. tig. 1) in its 

 district ; and it is the condition of all those numerous species of dillerent orders which 

 now appear fixed and distinct. "When (as happens at St. Paulo, where a greater abund- 

 ance of individuals and species, both of lliioinia and LcplaliH, exists than in the locality 

 of the last-named) many species have been in course of formation out of the varieties of 

 one only, occasional intercrossing may have taken ])lace ; this would retard the jiroccss 

 of segregation of tlH> sjieeies, and, in fact, aid in producing the state of things (varieties 

 ami half-form(>d species) which I have already described as there existing. 



In what way our Lfjildli-s originally acMpiired the general foi'in ami coloui's of ////owm' 

 1 must leave undiscussed. AVe may comdude (if we are to reason at all from existing 

 facts) that, as the antecedent forms of our races of Lcfifdlis which are still undergoing 

 (diange were themselves similai- to Uhoutia', the form has been inherited through a long 

 line of ancestors, which have been more or less subjected to similar conditions. The 

 instance of one of oui- forms leaving the Illiomia' to mimic a species of another family 

 may show us how a new line of mimetii- analogy and gradual modil'ication may have 

 been originally opened. 



Such, 1 conceive, is the oidy way in \\hich the origin oi' minu'tic speei(-s can be ex- 

 f)lained. I bcdieve the case offers a most Ixnuitiful ])i'oof of the truth of the theory of 

 natural selection. It also shows that a lU'W a(Ia])lation, or the fornuit ion of a new species, 

 is not elleeted by i;'i-eat and sudden change, but by nuincrous small sle])>. of natural 

 variation ami selection. Some (jf the nuitual resemblances ol' the //^'//ro/z/'Ar already 

 mentioned seem not to be due to the adaj:)tation of the one to the ot her, but rather, as they 



:•. z -1 



