HEREDITY 477 



mistaken for distinct species. In other words, their variations 

 are not continuous but discontinuous. 



Thus the earwig is of two distinct types as to size (dimorphism), 

 and many insects exist in three different forms, larva, pupa, 

 and imago, the crawling or worm form, the resting stage, and 

 the winged form. 1 



Sex in general means dimorphism, for, almost invariably, 

 marked differences exist between males and females of all 

 species. Sometimes, as in most mammals, the males are the 

 larger, but often the opposite is true, as in the case of many 

 birds and insects. External differences other than size, how- 

 ever, are certain to distinguish the sex by a number of non-sexual 

 characters. 



Dimorphism in improved breeds. Most of our improved breeds 

 exhibit more than one type entirely aside from considerations 

 of sex. For example, the Hereford is remarkably constant in 

 color, but there are two distinct types as to form. One is heavily 

 built and long-bodied, with deep flanks and straight thighs ; the 

 other is smaller and shorter, with less depth behind and a 

 tendency to rounded buttocks. The fore quarters are not differ- 

 ent in the two types, but the differences behind are marked 

 and the types do not readily blend. The breed appears to be 

 almost, if not entirely, dimorphic. 



Among the Shorthorns we have no less than half a dozen 

 types that do not readily mix. The pure white is distinct in con- 

 formation, as is the Duchess roan, the Cruickshank roan, the 

 cherry red, the dark mahogany red, and the Cruickshank red. 



The Percheron horse is dimorphic both as to color and as to 

 form. Whether it will always remain so, or will finally blend 

 into a common type as to color and conformation, time only 

 will tell. The same is true of the Jersey and the Holstein- 

 Friesian cattle, the Berkshire hogs, and the Shropshire sheep. 



All widespread and most newly developed breeds are polymor- 

 phic, the first from the external influences and different stand- 

 ards of selection, the second from recently associated dissimilar 



1 Excellent material on seasonal dimorphism of butterflies may be found in 

 Weismann, Studies on the Theory of Descent, I, i-ioo; and on polymorphism in 

 insects, Ibid. II, 401-481. 



