170 Inheritance in the Stock (Matthiola incana) 



be very much narrower with a straight or even concave edge as though 

 growth had been arrested. Where patches of hairs were so distributed 

 as to involve the leaf margin the limit of these areas would be defined 

 by a break in the even contour of the leaf outline, indicating a want of 

 conformity in growth. It therefore seems clear that some difference 

 more profound than the mere localisation of some substance directly 

 causing hair production and one having a marked effect upon growth 

 is here in question. 



Another aspect of the question is presented by the following class of 

 facts. We have seen that as regards the character hairiness in the Stock 

 a distinct genotypic maximum can be observed characteristic of the 

 different factorial combinations, but that this maximum is not attained 

 under all circumstances. The nature of the fluctuations suggests that 

 where the genotypic effect is normally slight some change in physiological 

 equilibrium precluding genotypic expression may be brought about by 

 causes unconnected with genotypic constitution. Such equilibrium we 

 imagine to be not a fixed and unvarying condition persisting for any 

 considerable period, but a perpetual balancing of various interdependent 

 and compensating activities resulting in constant, if slight, oscillations 

 about a physiological mean. If a character, regarded as arising in this 

 way, is strictly inherited such oscillations, due to inherent physiological 

 instability, are presumably too slight to affect expression of this genetic 

 character. But we can suppose that these sub-minimal oscillations 

 may become increased by environmental inequalities, and even reach a 

 point which renders somatic expression a physiological impossibility. 

 Or, it may be that the largest deviation so caused in the direction that 

 would lead to physiological inhibition still does not overstep the physio- 

 logical limit for somatic expression, in which case the genetic forces 

 (factors), with this margin to spare, have still full play and hence 

 produce their maximum effect. The manner in which the hairs are 

 distributed in certain grades points to the conclusion that some con- 

 nection, however indirect, exists between the hydrostatic conditions set 

 up within the plant, possibly at some critical stage, and the production 

 of hairs. This is most apparent in those forms in which hair formation 

 is so much diminished as to approach vanishing point, as for example 

 in the grade where the leaf shows no hairs except over the hydathodes. 

 If we accept this view it is easy to see how such irregularities as the 

 occasional absence of the terminal hair from a leaf in a trichome-bearing 

 series may be due to some such external cause, e.g. the dryness of 

 the soil acting either directly or indirectly through its effect on a 



