The Cytoplasm as Specific Substrate 209 



went in exactly the direction expected if the sperm plasm influenced 

 the egg cytoplasm (many curves and tables in 1924). This result 

 strongly favors the mitochondria as the suspected cytoplasmic agent. 

 Since mitochondria are the seat of cytochrome oxidases, a racial 

 difference affecting the kinetics of melanin production sounds reason- 

 able. We shall see later that in plants, also, comparable though not 

 identical differences between races were found by the Michaelis 

 group. 



b. Genomes in different cytoplasm 



In the cases discussed thus far the substrate action of the cyto- 

 plasm in influencing quantitative features of genically controlled 

 developmental processes was clear, because the effect could be 

 analyzed in simple Mendehan segregations. This applies to the cases 

 reported just now as well as to the material involving alternative 

 norm of reaction. However, in many cases analyzed in plants, the 

 cytoplasmic effect is a generalized one: it cannot be attributed to 

 interference with individual known genie actions, but rather inter- 

 feres with the whole genome, or major parts of it, if the genome is 

 located in the wrong cytoplasm. The diverse cytoplasmic actions may 

 be characterized, in a general way, as a weakening or inhibition of 

 general processes of growth, as a consequence of nuclei working in 

 the wrong cytoplasm. The technique, introduced by Correns (1908), 

 is as follows. Reciprocal crosses giving matroclinous phenotypes are 

 the starting point. These are backcrossed over generations to the 

 paternal form, with the result that the maternal cytoplasm remains 

 but the chromosomes are finally replaced more or less completely by 

 those of the paternal form, which are supposed not to fit into the 

 foreign cytoplasm. In a general way, the results are always the same: 

 continuation of the maternal effects over all generations without visi- 

 ble influence of the replaced genome or, as we shall see, with very 

 httle influence of it. 



It is remarkable that practically all the manifold effects found 

 are simple inhibitions. Many are in the sexual sphere, like suppression 

 of the male phase in monoecious plants ( gynodioecism ) , as in 

 Correns' work on Cirsium; or degeneration of male gametes at 

 different stages (pollen sterility) in Rhoades' work on maize (1933); 

 or the work of Lehmann and Michaelis and their students on crosses 

 of species and geographic varieties with backcrosses up to twenty- 

 five generations, which followed the pioneer work of Renner and 

 Kupper (1921). (See review by Michaelis, 1954.) It is characteristic 



