the Coloration of Eggs. 547 



considerable proportion of tliem to do so. Nor, again, 

 is it suggested that either colour, smell, or (necessarily) 

 taste are in themselves protective. The real protection, 

 causing occasional or frequent refusal (according to its 

 strength), is believed to be indigestibility in varying degree, 

 which can be overcome by sufficient gastric activity such as 

 is present especially in an empty stomach, and greater in a 

 half-filled stomach than a full one. Coloration and smell 

 would be of use merely to enable the enemy to recognise, 

 without breaking it, an egg, already known to him, that 

 he is not at the moment hungry enough to digest, and the 

 sight or smell of which, therefore, is sufficient to produce in 

 him a feeling of disinclination. 



Mimicry is worth defining briefly again in the sense in 

 which I shall refer to it below — not that I lay excessive 

 stress on its occurrence. It is the resemblance (brought 

 about by the selection of the appropriate variations, large 

 or small, "mutations'' or ''fluctuations") of an abundant 

 and more or less deterrent species (the "model") by a 

 species more liable to attack (the mimic). The latter's 

 greater liability to attack may result from a smaller power 

 to deter (whether by indigestibility or otherwise), or from 

 being less well known to enemies, and so more liable to 

 suffer from mistaken attack ; or from a combination of these 

 relative disadvantages. It is even conceivable that points of 

 resemblance occurring as between two equally deterrent and 

 al)undant species might be selected owing to the doubled 

 reminding-power afforded by the combined populations and 

 the facilitated task for the memory provided by recognition 

 characters in common. A "mimetic association" is the 

 colour-group formed by a model (more or less indigestible 

 or otherwise deterrent) and its various mimics, whether 

 these have mimicked it for increased notoriety* (being 



* Not really Miillerian mimicry, though the basis is still mainly 

 population. Miiller s view was not " more reminders, simplified recog- 

 nition, and less mistaken attack," but "equal destruction (by young 

 enemies) whatever the population," the greater population thus losing a 

 smaller percentage. The first view seems the more probable and is 

 supported by my experiments. 



2 p 2 



