XV] Heredity and Repetition of Parts 275 



changes and developments, while remaining attached to 

 the other half, we call the resemblance between the two 

 halves Symmetry ; but if the division is one by which two 

 new individuals are formed, and the two halves separate 

 and lead independent lives, then we ascribe the resemblance 

 between the two individuals to heredity. On a previous 

 occasion (11) I pointed out that when the comparison is 

 followed in the other direction it appears that if 



I, Symmetrical Repetition of Parts is comparable 

 with Heredity, 



then 2. Differentiation between Parts is comparable with 



Variation, 



That there must be limits beyond which the comparison 

 fails, is clear enough, but I do not think they have been 

 yet satisfactorily defined. The discovery of a true delimi- 

 tation of the properties and attributes of individuals, which 

 distinguish them from parts, would constitute a great advance 

 in biological theory. Perhaps the nearest we can get to 

 such a distinction is a recognition of the fact that ordinary 

 somatic differentiation is ^/i"?^^//)/ distributed in a symmetrical 

 manner about one or more axes, while among gametic 

 tissues such axes are not tcstially perceptible ; but to both 

 statements there are some notable exceptions. Bud-sports, 

 however, never, so far as I know, are distributed symmetri- 

 cally about the axis of the individual producing them, and 

 thus are distinguishable from ordinary somatic differentia- 

 tions. Dr S. F. Harmer in his Presidential Address to 

 Sect. D of the British Association (Dublin, 1908), describing 

 the various forms of avicularla"^ found on the same colony 

 of certain Polyzoa, made the interesting suggestion that 

 there may be an allelomorphic relationship between these 

 parts. Sometimes one special type of avicularia characterises 

 a species, sometimes another type ; while again both types 

 may occur together on the same colony. Now this is a 

 case to which such a suggestion is especially applicable ; 

 for the different types of avicularia are not distributed In 

 a symmetrical pattern, but apparently at random on the 



* These structures are prehensile organs of various patterns, somewhat 

 resembhng minute crab's claws or the heads of birds. Morphologically 

 they are regarded as much modified " individuals." 



18—2 



