210 Cytoplasm as Seat of Genetic Properties 



that the disturbances produced by the genome-plasmon incompati- 

 bihty are lethaHty, steriHty, abnormahties of stem, leaves, and flowers, 

 cripples, dwarfs, loss of anthocyanin, and also heterosis. In addition, 

 degrees of cytoplasmic viscosity and permeability, sensitivity to 

 poisons and fungi, different enzymatic activity, and different general 

 physiological reaction? were found. Michaelis reports that nearly 

 all properties of the plant are altered by cytoplasmic differences. 

 This variability is not much lower than that which would be pro- 

 duced by a series of gene mutations. Michaelis concludes that this 

 is definite proof that the cytoplasm contains autonomous genetic con- 

 stituents. 



I cannot concur in this conclusion. The type of effects studied 

 suggests rather a completely unspecific cytoplasmic action, which we 

 described in terms of substrate action. A good model for it is a dif- 

 ferent constitution in regard to the oxidation-reduction system, lo- 

 calized within the mitochondria, which works successfully only with 

 the proper products of the genie material; in the wrong combination, 

 therefore, it probably upsets basic reactions in development only 

 in regard to their kinetics, with the result of widespread defects. In 

 more descriptive than correct language we might say that the wrong 

 cytoplasm "poisons" developmental reactions. 



The problem is so important that we may repeat our conclusion 

 in other words. It is safe tc say that in this whole group of facts 

 cytoplasmic heredity in a strict sense is not involved. The cytoplasm 

 or plasmon does not control the development of definite hereditary 

 traits. It is rather a substrate for the action of the genie material, 

 and it functions properly only when it has the proper chemical com- 

 position needed to react correctly with the products of genie action. 

 The chemical composition may differ in somewhat different species or 

 geographic varieties; it may be extreme; or it may be completely 

 absent. The difference may affect only the quantitative features of 

 certain gene-controlled processes as in Lymantria. Or it may affect 

 only the all-or-none reactions involved in the decision of the sexual 

 alternative. (See above the parallel facts in the chapter on maternal 

 inheritance.) Or it may "poison" many of the gene-controlled re- 

 actions involved in the general processes of growth and differentia- 

 tion. The part of the cytoplasm that has to do with the control of 

 oxidase action, as in mitochondria, would fulfill most if not all the 

 requirements on which the cytoplasmic action under discussion is 

 based. 



This conclusion is considerably strengthened by experiments of 



