198 THE io\(iLiirioN 'imikody 



niiu'o at ovory roclm'iiiL:; di\isuiii tlio itls ol' (lie ^t'rin-fcll.s nr(> liroui^'hti 

 down to hall' tlioir nunilu'r, a possibility is otten-d for ^■radimily 

 roiuoving- the md'aN-oiiraMo ids IVoin the o-erm-plasiw ol' the sin'cii-s, 

 since thodcseonda.nts resulting- I'roiu (ho luosi uitrn\(UH'alilo id-comliiua- 

 tiions always povish, and so IVou> oTncratiou to j;-onera(ion tho i;vrni- 

 plasni o'ots ridol' its \inl"avouval>ly varying' ids, and the most propitions 

 combinations afforded by amphimixis :\\v pr(>served, till uHimaldy 

 there remain only those combinations which are \ aryin_i;' appropriately, 

 or at least inily those in which the a]>proiiri,'U.ely wn-yinn' de(.ern\ina.nts 

 are in the majority, and so have eontrollinij,' intluence. 



Log'ically this dednction is undoubtedly indispn(al)l(\ from the 

 standpoint of the i;-eriii-pla.sm theory : but whether it may be r(>i;'ardod 

 as a sufficient reason for the introduction ol' amphimixis, and I'oi- 

 its exti'emiily tenacious ])ersistence (.liroun'hout (lu' course ol' (ho 

 loiio- and intricate phyloijeny, cannot be maintained witho\i(. special 

 investin'atiou. 



Ao-ainst my position the objection has often Ihh'u nru'ed thai- an 

 arran^enu'ut cannot arise or be maintained thron!j;h natural st'leotion 

 unless it is of illirrt iinc to tho individual in which \l occurs. Sexuail 

 reproduction cannot therel'ore have been established simply because 

 it adx'ances, or ex'en because it makes possible the ada.p(.alions of 

 species, for these adaptations only ca.me about occasionally, perhapiH 

 once in a thousand };'imera.tions or even less I'nHpientJy ; thus the 

 inter\'einnn' generations could derix'e no adxanta^j'e of any Icind from 

 tlie arranu'ement in question, and therel'ore, accordini;' to the law of 

 the dcf^'eneration of unused characters, it nnist ha-xe lonj;' sinci^ been lost. 

 I mention(>d this objection bel'ori', but wa,s obliti'ed to postpone 

 a detailed eonsidera-tion of it until we had discussed freruuna,! 

 selection. 



Wo admit, of course, tliat charact-ers are only preserved int-act 

 as long aw tlwy are of advantages siiffici(-nt lid turn tlie sca-le in favour 

 of their ])ossesHors, and that they begin to fall from their height of 

 perfection Avhen that is no longer the case; we ailnut also that lunv 

 a(la])tations are not continually' necsessarry, but- are so only at interva-is 

 of long series of generations, a-ud yot the* objection cited seems t-o me 

 baseless. 



Leaving out of account, for tlu^ moment, t-la^ iirst introduction of 

 amphimixis, let ns <leaJ with it as an existing occuri'ence, for tlie 

 tenacious persisteiicis of which we wish to lind rea-sons. 



Is it really tho case that amphimixis is only of iinportaiUHi in 

 connexion with tho new adaptation (if a s]iiH',ies, and tliat it- has not-hing 

 to do with the persisteiuse of the spiM'ies in this sta-to of adapta-tion 



Digitized by Microsoft® 



