II. MLSSON-KIIU:. — MKNDÉLIS.MK ET ACCLIMATATION. 157 



'Jnil.— If llie parents dilTer con.çiderably l'rom eacli olher as regards precocily, 

 llieii a large majorily of iiilermediale forms are obtained whicli may beconie 

 conslaiit, whereas Ihose forms which resemble Ihe parents, or those which are 

 earlier or laler Ihan eilher parent, are relatively rare. In some cases there is a 

 majority of forms resembling, or nearly resemljjing, one of the parents, for 

 instance, the laler forms may prevail. 



ôrd. — In the case of crosses between forms of the same precocily, ordiffe- 

 ring slightly from each olher, forms may arise which are earlier or laler Ihan 

 eilher parent. In some of Ihese cases Ihe dilïerences are very slighl, biil in 

 olhers Ihey are definite and considérable The extrêmes are connecled wilh 

 the parental types by transitional forms. The origin of thèse new types cannot 

 be due lo sponlaneous changes or to any accidentai cause, but are solely the 

 resull of ségrégation. Siniilar transgressive forms are prodiiced, through con- 

 tinued ségrégation, in the following générations. 



4th. — By hybridizalion of constant types, resembling each olher or difTering 

 only slightly in precocily, a range of variation may be produced extending over 

 a grcat pari of the spécifie range of variation for tliis characlcr, and extrême 

 forms may occur which appear to exceed the limits hithcrto known. 



My observations on the characler of résistance to cold inautumn wlioat are 

 less numerous, but the rcsults are analogous. 



AU observations hilherto made on ségrégation are in agreemenl with the 

 idea thaï thèse results are due lo the séparation of several mendelian faclors 

 which by varions recombinations give rise to a long séries of forms. 



Thèse forms are therefore nol mutations, but represent variouscombinations 

 of a relatively small number of faclors. In the cereals, in which self ferti- 

 lization is the rule, the true breeding forms must be regarded as homozygous 

 with regard to thèse faclors. Il is especially lo be observed thaï similar, or 

 very nearly similar forms, may l)e (luile dilTcrenl combinations of faclors, as is 

 shown by the facl thaï from a cross between Iwo such similar forms there may 

 arise new types in which certain quantitative characters exceed. sometimeseven 

 in a high degree, the range of thèse characters found in the parents. The 

 same externat appearencemay be Ihe manifestation of diverse internai constitu- 

 tion, or, in olher words, may be produced by combinations of dilïerent faclors. 

 The sélection, according to Iheir external appearance and not according lo 

 their gamelic constitution, of types besl adapted to their conditions, is thus 

 only one aspeclof acclimatization. Inasmuch as thèse selecled forms represent 

 dilTerenl factorial combinations, there exists the possibilily of producing by hy- 

 ])ridi/.ation new combinations which may be still bélier adapted, and Ihis me- 

 lliod of improvemenl may be continued lill the mosl favouralde combination is 

 found. Acclimatization, or adaptation, means then, from this point of view, 

 a regrouping of the components or mendelian faclors already existing, giving 

 more advanlageous combinations, 



Withoutdenying Ihe possibilily ofollier modes ofacclinializalion, thaï \\liirli 

 I hâve described, and which occurs by the recombination of faclors already 

 existing, is Ihe only method Ihal I hâve been able to verify through experimenls 

 wilh the cereals. Il must be confessed however thaï this mode of acclimatiza- 

 tion is the niost easy one which may be staled by empirical investigations. 



IIow are mendelian faclors produced? Thatis the problem which now pré- 

 sents itself, and from the true solution of which we are probably still far distant. 



