THE ALTERNATION OF GENERATIONS. 35 



nutriment. But it is a living nutriment, in some respects 

 comparable to that which would be supplied to an animal 

 kept alive by transfusion, and its molecules transfer to the 

 impregnated embryo cell all the special characters of the or- 

 ganism to which it belonged. 



The tendency of the germ to reproduce the characters of 

 its immediate parents, combined, in the case of sexual genera- 

 tion, with the tendency to reproduce the characters of the 

 male, is the source of the singular phenomena of hereditary 

 transmission. No structural modification is so slight, and no 

 functional peculiarity is so insignificant in either parent, that 

 it may not make its appearance in the offspring. But the 

 transmission of parental peculiarities depends greatly upon 

 the manner in which they have been acquired. Such as have 

 arisen naturally, and have been hereditary through many an- 

 tecedent generations, tend to appear in the progeny with 

 great force ; while artificial modifications — such, for example, 

 as result from mutilation — are rarely, if ever, transmitted. 

 Circumcision through innumerable ancestral generations does 

 not appear to have reduced that rite to a mere formality, as 

 it should have done if the abbreviated prepuce had become 

 hereditary in the descendants of Abraham ; Avhile modern 

 lambs are born with long tails, notwithstanding the long-con- 

 tinued practice of cutting those of every generation short. 

 And it remains to be seen whetlier the supposed hereditary 

 transmission of the habit of retrieving among dogs is really 

 what it seems at first sight to be ; on the other side, Brown- 

 Sequard's case of the transmission of artificially-induced epi- 

 lepsy in Guinea-pigs is undoubtedly very weighty. 



Although the germ always tends to reproduce, directly or 

 indirectly, the organism from which it is derived, the result 

 of its development difi'ers somewhat from the parent. Usually 

 the amount of variation is insignificant ; but it may be con- 

 siderable, as in the so-called " sports ; " and such variations, 

 whether useful or useless, may be transmitted with great te- 

 nacity to the offspring of the subjects of them. 



In many plants and animals which multiply both asexually 

 and sexually there is no definite relation between the aga- 

 mogenetic and the gamogenetic phenomena. The organism 

 may multiply asexually before, or after, or concurrently with, 

 the occurrence of sexual generation. 



But in a great many of the lower organisms, both animal 

 and vegetable, the organism (A) which results from the im- 

 pregnated germ produces offspring only agamogenetically. 



