October 21, 1920] 



NATURE 



257 



set of chromosomes, tach icarrying a complete set 

 of the factor elements. Hence, if some of the one 

 set become eliminated, we can still imaj<ine that a 

 normal, though undersized, individual might develop. 

 The converse relation, where increas<'d size goes with 

 multiplication of chromosomes, was discovered by 

 Gregory (Proc. Roy. Soc., B, vol. l.xxxvii., 1914) in a 

 Primula, and occurs also in Ginothera gigas, a 

 mutant derived from CE. Laniarckiaua. It is in- 

 teresting in this connection to recall the results ob- 

 tained by Nemec {Jahrb. /. wiss. Bot., xxxix., 1904, 

 •Das Problem der Begruchtungsvorgange," 1910) as 

 the result of subjecting the root-tips of various plants 

 to the narcotising action of chloral hydrate. Under 

 this treatment cells undergoing division at the time 

 were able to form the daughter nuclei, but the pro- 

 duction of a new cell-wall was inhibited. The cells 

 thus became binucleate. If on recovery these cells 

 were to fuse before proceeding to divide afresh a 

 genuine tetraploid condition would result. So few 

 cases of natural tetraploidy have thus far been ob- 

 served that we have as yet no clue to the cause which 

 leads to this condition. 



The conclusions to which we are led by the con- 

 siderations which have here been put forward are, in 

 the main, that we have no warrant in the evidence 

 so far available for attributing special hereditary pro- 

 cesses to the cytoplasm as distinct from the nucleus. 

 On the other hand, there is a very large body of 

 facts pointing to a direct connection between pheno- 

 typic appearance and chromosomal l>ehaviour. In 

 animals the evidence that the chromosomes constitute 

 the distributional mechanism may Ix- looked upon 

 as almost tantamount to proof ; in plants the observa- 

 tions on Droscra, Primul;i, Ouiothera, and Sphaero- 

 carpus are in harmony with this view. When we 

 ( omc, however, to the question of linkage and general 

 .-ipplicability of the conception of "crossing-over" as 

 .idopted by Morgan and his school, we are on less 

 (f-rtain ground. In Drosophila itself, the case which 

 the scheme was framed to fit, the entire absence of 

 " crossing-over " in the male remains unaccounted 

 for, while the evidence from certain plant tvfK-s 

 .-ippears to l>e definitely at variance with one of its 

 fundamental premises. If segregation at the rccng- 

 nised reduction division is definitely established for 

 animal types, then we must conclude that the sorting- 

 out process may follow a different course in the 

 plant. 



The question as to what is the precise nature of the 

 differences for which the Mendelian factors stand is 

 constantlv before the mind of the breeder, but we 

 are only now on the threshold of investigation in this 

 direction, and it is doubtful wheth<r we can as yet 

 give a certain answer in anv single instance. Still 

 less are we able to say what the actual elements or 

 units which undergo segregation may be. In the 

 case of such allelomorphic pairs as purple and red 

 snp colour or while or cream plastid colour, it may 

 Yx- that the difference is wholly qtinlitalive. consisting 

 merelv in the formation or non-formation of some one 

 chemical substance. But the majority of charac- 

 teristics are not of this hard-and-fast tyjie. Between 

 some the distinction appears to be one of range- -lo 

 be quantitative rather than, or ns well as, qualitative 

 in nature, ami range must mean, presumably, either 

 cumulative effect or a force or rate difference. It 

 mav well be, for example, tlvit with some change in 

 phvsiological equilibrium accompanying growth and 

 development, factorial action mav be enhanced or 

 accelerated, or, on the other hand, retarded or even 

 inhibited altogether, and a regional gradinp result in 

 consequence. Range in a < haracter is not confined 

 to, though a common characteristic of, individuals of 



NO. 2660, VOL, 106] 



cross-bred origin. It may be a specific feature, both 

 constant and definite in nature. For example, a 

 change as development proceeds from a glabrous or 

 nearly glabrous to a hairy condition is not an un- 

 usual occurrence in plants. In the stock such a 

 gradational assumption of hairiness is apparent no 

 less in the homozygous form containing a certain 

 weak allelomorph controlling surface character, when 

 present with the factors for sap colour, than in those 

 heterozygous for this or some other essential com- 

 ponent. We see a similar transition in several 

 members of the Scrophulariaceae, e.g. in various 

 species of Digitalis, in Antirrhinum niajiis, A. oron- 

 tiiim, Anarrhinttm pedatttm, Pentstemon, and 

 Nemesia. In perennials an annual recurrence of this 

 change of phase may be seen, as in various species 

 of Viola and in Spiraea ulniiaria. In some, perhaps 

 in all, of these cases the allelomorphs mav stand for 

 certain states of physiological equilibriurli, or such 

 states may be an accompanying feature of factorial 

 action. .\ change of phase may mean .in altered 

 balance, a difference of rhythm in interdependent phvsio- 

 logical processes. In the case, for instance, of a cer- 

 tain sub-glabrous strain of stock in which the pre.sence 

 of a single characteristically branched hair or hair- 

 tuft over the water-gland terminating the midrib in 

 a leaf otherwise glabrous is an hereditary character, 

 it is scarcely conceivable that there is a localisatioti 

 in this region of a special hair-forming substance. It 

 seems more probable that some physiological condition 

 intimately connected with the condition of water- 

 content at some critical period is a causal factor in 

 hair production, and that this condition is set up over 

 the whole leaf in the type, but in the jwrticular strain 

 in question is maintained only at the point which 

 receives the largest and most direct supply. In this 

 same strain a leaf may now and again be found 

 lacking this hydathode trichome in an otherwise con- 

 tinuous hair-forming .scries, an occurrence which mav 

 well result from a slight fluctuation in physiological 

 equilibrium such as is inherent in all vital processes — 

 a fluctuation which, when the genetic indicator is set 

 so near to the zero point, may well send it off the 

 scale altogether. If, as is not improbable in this and 

 similar cases, we are concerned with a complex chain 

 of physiological processes, investigation of the nature 

 of the differences for which the allelomorphs stand 

 may present a more difllcult problem than where the 

 production of a particular chemical compound appears 

 to be involved. In such a physiological conception w-e 

 have probably the explanation of the non-appearance 

 of the recessive character in certain dominant cross- 

 breds. 



Up to this point we have treated of the organi.sm 

 from the aspect of its being a wholly self-controlled, 

 independent system. .\s regards some characteristics, 

 this may be regarded as substantially the case— that 

 is to say, the soma reflects under all observed condi- 

 tions the genetic constitution expressed in the Men- 

 delian formula. Correspondence is precise between 

 genotypic potentiality and phenotypic reality, and we 

 have so far solved our problem that we can prinlict 

 certainly and accuratelv the appearance of offspring, 

 knowing the constitution of the parents. In such 

 cases we mav say that the efficiency of the genetic 

 machine works out at 100 per cent., the influence of 

 external environment at o. Our equation somatic 

 appearance = factorial constitution requires no correc- 

 tion for effect of conditions of temper.itiire. humiditv, 

 illumination, and the like. But most somatic 

 characters show some decree of varinbllitv. Pheno. 

 fvpic appearance is the outcome primarilv of genotNTjic 

 constitution, but upon this are superposed fluctuation*, 

 slight or more pronounced, arising ■ ' • -ulf of 



