6. Why some species are more variable than others. 7. Why certain variations tend to occur very frequently. 8. Why variations in some directions seem never to occur. 9. Why a female may produce offspring when paired with one male of her species and not when paired with another male of the species. 10. Why organisms that arise by parthenogenesis appear to be as variable as those which are sexually produced. 11. Why certain animals possess the power of regenerating lost parts, while others have not this power. 12. Why most plants and some of the lower animals can be produced asexually from cuttings. 13. Why mutilations are not inherited. 14. Why acquired characters are rarely, if ever, inherited. 15. Why the ovum puts forth the polar bodies. 16. Why the mother-cell of the spermatozoa produces four spermatozoa. 17. Why differences in the nature of the food administered to the larvæ of ants determines whether these shall develop into sexual or neuter forms. 18. Why the application of heat, cold, etc., to certain larvæ affects the nature of the imago, or perfect insect, to which they will give rise. 19. Why the females in some species lay eggs which can produce young without being fertilised. 20. Why some species exhibit the phenomena of sexual dimorphism, while others do not. 21. In addition to all the above, a satisfactory theory of inheritance must account for all the varied phenomena which are associated with the name of Mendel. It must explain the various facts with which we have dealt in the chapter on hybridism, why some species produce sterile hybrids when intercrossed, while others give rise to fertile hybrids, and yet others form no offspring when crossed; why the hinny differs in appearance from the mule, etc. 22. It must explain all the facts which constitute what is known as atavism. 23. It must account for the phenomenon of prepotency. 24. It must explain the why and the wherefore of correlation. 25. It must tell us the meaning of the results of the experiments of Driesch, Roux, and others. 26. It must render intelligible the effects of castration on animals. Existing Theories Unsatisfactory Now, no existing theory of heredity can give anything approaching a satisfactory explanation of all these phenomena. It is for this reason that we refrain from critically examining, or even naming, any of them. We are convinced that in the present state of our knowledge it is not possible to formulate anything more than a provisional hypothesis. It must not be thought that we consider the various theories that have been enunciated to be of no value. Erroneous hypotheses are often of the greatest utility to science, for they set men thinking and suggest experiments by means of which important additions to knowledge are made. We now propose to set forth certain facts of inheritance, and from these to make a few deductions—deductions which seem to be forced upon us.