Problems of Gametic Constitution 79 



isms are composed of ultimate " biological units of structure " which must be 

 thought of in the same manner as Weismann conceived of his biophores. Any 

 such idea is essentially vitalistic, for the reason that the unit is in all respects 

 a living individual, and between it and the non-living there must be a sharp 

 division, or else we must think of more remote agents of composition, and so on 

 without limit. 



On the other hand, if we think of the observed results produced in the soma 

 merely as products of gametic reaction centers of a purely physical nature, no 

 difficulty is introduced into the interpretation of the observations, because there 

 are in physical phenomena ample evidences of the fragmentation of reaction foci 

 into two or more discontinuous and independently acting centers, which may 

 be permanent or transient, depending entirely upon the surrounding materials 

 and conditions in the medium. 



The basis of the ideas with regard to the fixity of these ultimate units of 

 structure and activity was derived first from the interpretation of nature and 

 evolution phenomena, and later by the neo-Mendelian experiences in crossing, 

 in which morphological considerations dominated the logic and philosophy of 

 the writers, developing dogmatic definitions and concepts of the discrete, dis- 

 continuous nature of both agents and of somatic characters. The neo-Mendelian 

 observations showing the alternative transfer of characteristics is no criterion 

 for the deduction that the agents are discrete and exist as units. It is con- 

 venient to symbolize them for purposes of description, but it must not be for- 

 gotten that it is only a symbolization and nothing more. In organisms we are 

 not dealing with static masses, but with highly complicated integrations of 

 material, in which endless physical and chemical reactions are in ceaseless 

 operation. In this mass there are innumerable centers of activity at all times, 

 in the germ as in the soma, and as the result of experiences with the materials 

 that I have had to study, I can only regard the gametic agents as expressions of 

 centers of activity of greater or less magnitude, bound into a system of opera- 

 tion w^hich in the main retains its totality of manifestation throughout successive 

 generations ; but this retention of uniform aspect is not so much the product 

 of its structure and composition as it is that it has in its successive generation 

 cycles the same kind of materials to react with, and also essentially the same 

 conditions in the medium in which to carry on these reactions ; therefore repeti- 

 tion of reactions and results must naturally follow. 



The fact of the perceptual experience of the metathesis of characters in cross- 

 ing seems to me no evidence of the transfer of fixed units of any kind. The 

 same gametic system, L. signaticollis, in one combination acts as a unit, in 

 another combination it acts as two, or three, or more, depending upon the losses 

 or replacements in its minor portions. Moreover, in my materials I have not 

 been able to decide where the lower limit of unit manifestation was to be drawn 

 in any instance. For example, in the complex pronotal pattern it may behave 

 in crossing as an entity and may be symbolized as a unit of activity, or specific 

 and trivial elements of it may show this same capacity for metathetic action, 

 giving typical ]\Iendelian reactions, and discontinuity of the somatic manifesta- 

 tions, and stable extracted products, in F,, and in no instance have I been able to 

 satisfy myself from experimental evidence that I had attained the limit, present 

 in the materials for the production of this alternative, discontinuous type of 



