266 MR. E. B. POULTON ON THE PROTECTIVE [Mar. 1, 



suggests that the enemies are startled by the sudden 'manner in 

 which the bright colours are displayed ; but I am not aware that a 

 similar significance has ever been attributed to the bright colours 

 of Butterflies suddenly seen when the wings are opened. The former 

 suggestion probably holds, for I think the margin of the underwings 

 is more commonly found to be notched than any other part of these 

 insects when ca])tured. But the primary significance of such bright 

 colours, concealed in the protective attitude of rest, must be the 

 same as those of Butterflies, and I should attribute the same meaning 

 — of sexual adornment — to the brilliant colours of the underwings 

 of the Grasshoppers of the genus (Edipoda, also alluded to by Jenuer 

 Weir (see Trans. Ent. Soc. Lond. 1869, pt. i. p. 23). 



Conclusions. 



The following are the general conclusions arrived at by the con- 

 sideration of the experimental data tabulated in this paper : — 



1. The extremely speciahzed defence of the larval stage follows 

 from its delicate anatomical construction and the necessities which 

 are imposed on it as the great feeding-stage. 



2. Highly conspicuous insects nearly always possess some un- 

 pleasant attribute, i. e. a disagreeable taste or smell in the tissues 

 and fluids of the body, or (in the case of the smell) discharged from 

 special glands ; irritating hairs ; or stings. 



3. The conspicuous appearance may be due to strongly-contrasted 

 colours, the presence of hairs or tufts, and the attitude in which the 

 body is held, and to gregarious habits, or attention may be 

 attracted by violent movements which take place when an enemy 

 appears. 



4. In a small number of cases a highly conspicuous appearance 

 has not yet been shown to be attended by any unpleasant attribute. 



5. In the various sjjecies in which a conspicuous appearance is 

 produced by colour and marking, the same colours and patterns 

 appear again and again repeated. In this way the Vertebrate 

 enemies are only compelled to learn a few types of appearance, and 

 the types themselves are of a kind which such enemies most easily 

 learn. Furthermore certain appearances are especially impressed on 

 the vertebrate foes by highly aggressive insects, feared because of 

 stings &c. ; and hence there is especial advantage in any approxi- 

 mation to such types. Again, the selected type of conspicuous 

 appearance also depends on the (probably protective) colours which 

 existed at the time when the conspicuous appearance first com- 

 menced (these can be determined with a great degree of probability 

 in some few cases). 



6. In a relatively few cases aggressive forms among the Vertebrata 

 (Serpents) are mimicked, although such an appearance is pure 

 intimidation, for the insect is quite harmless. 



/. It is not uncommon for an insect to be protectively coloured 

 but when detected to assume a terrifying attitude, and in some 

 cases to take up off"ensive measures (discharge of irritating fluid, &c.). 



