5o8 SCIENCE PROGRESS. 



generate representatives of the groups of four cone cells 

 each, which are present in normal retinulse. 



We have then in the case of the blind dwellers in 

 darkness a physical condition which is uniform and constant 

 throughout life, and a structural condition obviously related 

 to it which is not direct but recapitulative in its develop- 

 ment. The development of the eyes in these forms offers 

 a marked contrast therefore to the entire absence of the fore- 

 limbs of snakes at all ages, and resembles the development 

 of teeth in the gums of the foetal whale. The following 

 conclusions follow necessarily from the facts, 



(i) The blindness of cave animals has certainly not 

 been produced by the selection or survival of individuals in 

 which the eyes were defective from their first development. 

 If it is due to selection at all, it is the selection of indi- 

 viduals in which the eyes underwent progressive deteriora- 

 tion after the commencement of independent life. 



(2) The eyes even in the stage in which they are most 

 developed are far from being as well developed as in the 

 ancestors which lived in daylight, but are only somewhat 

 more developed than in the adult animal. Recapitulation 

 does not therefore occur except when the external condition 

 to which the ancestral structure was adapted continues to 

 act at an early period of life. 



In this case, therefore, as in many others, if we persist 

 in applying the selection theory of evolution we must be 

 content to infer that the variations occurred in particular 

 ways without any particular reason. If hereditary varia- 

 tions are independent of the external conditions of life there 

 is no reason why the eyes of blind animals should be better 

 developed in the young than in the adult. It cannot be 

 maintained that this is due to a general law of inferior 

 variations, that the absence of selection allowed individuals 

 with inferior eyes to survive and breed, and that congenital 

 inferiority of an organ always manifests itself by its de- 

 generation at a late period of life. On the contrary we 

 know plenty of cases in which an organ is absent in the 

 young and fully developed in the adult. The facts compel 

 us to conclude that the ontogeny of the individual has been 



