SEXUAL SELECTION. 83 



high wind to a small oceanic island and have successfully 

 established themselves there. Among the variations pro- 

 duced in the fresh generations some will have larger and 

 better-developed wings than others. These will run more 

 risk of being blown to sea and perishing, whilst their wings, 

 being no longer required for spreading the species nor 

 for protection against terrestrial enemies, become a 

 positive handicap in the search for food. In a few genera- 

 tions the variations with smaller wings will become pre- 

 dominant and eventually a wingless variety will be produced. 



Again, we obtain from the same principles a plausible 

 explanation of the extraordinary phenomena of Protective 

 Resemblance and Mimicry referred to in Chapter IX. 

 (See page 78.) 



An " accidental " variation causing an individual to bear 

 a faint resemblance to an inanimate object may be sufficient 

 to give it partial immunity from ever-watchful foes, and such 

 variations transmitted and accentuated may in time produce 

 these phenomena, which appear to imply such purposeful 

 resemblance. 



The student should be careful to recognise that Natural Selection 

 is only a step, however important, in the explanation of evolution. 

 Zoologists are still groping in the dark with respect to the origin 

 and transmission of variations and the factors determining heredity. 

 The most important question pressing for solution is — Does Natural 

 Selection work through the experimental method of selecting from a 

 number of indefinite variations, or are the variations produced in a 

 defiftite manner in response to the environmental needs ? The only way 

 in which the variations can be definitely related to the environmental 

 needs is as follows : — During the life of an organism, especially during 

 its early stages, it is susceptible to external impressions which leave an 

 indelible mark upon its adult structure. Two individuals with the 

 same parents and the same hereditary tendencies may be subjected to 

 environments so dissimilar that they become structurally adapted in 

 different directions. These adaptations are called acquired characters 

 (see Introduction). If we assume that the offspring of these individuals 

 have the acquired characters transmitted to them, even in a modified 

 degree, then the acquired characters of one generation become the 

 hereditary characters of the next and the adaptation in nature has a 

 simple explanation. This theory of evolution involving the Trans- 

 mission of Acquired Characters is connected with the name of Lamarck. 

 The transmission of acquired characters has never yet been experiment- 

 ally demonstrated and has been strenuously denied by Weismann and 

 others. Should such a process be indubitably proved to take place 

 in nature, natural selection would take a subordinate position as a 



