222 THE BEHAVIOUR OF POLYPLOIDS 



organs of the plant is found after doubling twice in Funaria hygro- 

 metrica (v. Wettstein, 1924), and after doubling only once in 

 A crodadium cuspidatum. The haploid is usually of a more immature 

 stripling growth than the diploid, while the tetraploid is of a stouter 

 growth (Jorgensen, 1928 ; Davis and Kulkarni, 1930 ; Clausen and 

 Mann, 1924). It should be noted that this applies to the fruits for 

 different reasons. The development of many-seeded fruits is often 

 proportional to their seed content, which is lower in haploids, 

 triploids and non-hybrid tetraploids than in the corresponding 

 more fertile diploids. Hence the fruits are smaller in these, and 

 often of different shapes (Blakeslee and Belling, Jorgensen). 



Organisms are therefore modified in general character by a 

 change in their size, i.e., by a change in volume-surface relations. 

 The same chromosome complement, in fact, always has a 

 qualitatively as well as a quantitatively different phenotypic 

 expression when represented different numbers of times, and 

 changes from the condition to which it is adapted, whether haploid, 

 diploid or tetraploid, are more or less deleterious, just as are changes 

 of balance. This principle is of importance in considering the 

 origin of the special differentiation found in species where the 

 male is haploid and the female diploid. In these the complement is 

 adapted to both the haploid and the diploid conditions. The 

 differentiation in phenotypic expression is then carried to an 

 extreme. It applies even to cell-size. Exceptional diploid males 

 in Habrobracon are no larger than the normal haploids (Torvik, 

 1 931). Special genes or gene-combinations can therefore be 

 selected to modify the ordinary relationship of the size of the 

 organism to that of the nucleus when the chromosome number is 

 altered, just as they are evidently selected to modify the sexual 

 character of the organism (Ch. IX). Simple doubling in species 

 not specialised in this exceptional way is usually accompanied by 

 more or less gigantism. Where the polyploid form shows no 

 increase in size, this can be ascribed to segregation of dwarfing 

 factors in the parent, e.g., in Crepis tectorum (Navashin, 1926), 

 where great size variation occurs in the diploid. It may be that 

 in some species giant polyploids are not fitted to survive, and the 

 failure of polyploidy to appear in many groups of flowering 



