The Cytoplasm as Specific Substrate 207 



bracon, also involving a pigmentation process, and carried it to one 

 more generation. 



The question arises whether a more detailed characterization 

 of the cytoplasmic effect in these experiments is possible. It is clear 

 that no effect of a qualitative type takes place. The affected pheno- 

 genetic process is the quantity of melanin deposited at certain points 

 of the cuticula, and this again is due to the speed with which melanin 

 is formed, as a statistical check upon the progress of pigment forma- 

 tion in all the instars of the larvae demonstrated. "Thus some kind of 

 generalized substrate action is involved in the cytoplasmic effect. This 

 does not necessarily mean a biochemical quaHty or a condition of 

 viscosity or colloidal behavior of the cytoplasm as an entity. Another 

 possibility might be a difference in formed constituents of the cyto- 

 plasm of a type which we have already met in the visible germ 

 track determining substances discussed previously. The most obvious 

 corpuscular elements are the mitochondria, which sometimes contain 

 a complete set of respiratory enzymes. These may be involved in 

 the speed and quantity of melanin formation. A few remarkable re- 

 sults point in this direction. According to Winge and Laustsen 

 (1940), a haploid ascospore in yeast may be fertilized and form 

 a diploid clone; or it may do so by autogamy (called direct diploi- 

 dization). The latter clones have a low viability, which is inherited. 

 For this the mitochondria are believed to be responsible. Winge and 

 Laustsen think that in autogamy the mitochondria have no time to 

 duplicate and, therefore, are present only in a quantity which suffices 

 for a haploid but not for a diploid cell. 



A very remarkable case, which probably is of the same type 

 as the Lymantria experiments, has been described by Oehlkers 

 (1952a), though it is a little more complicated. The hybrid Strepto- 

 carpus Rexii X Wendlandii is fertile, and the F2 and all backcrosses 

 can be obtained. The two parental forms have the same kind of sym- 

 petal flowers, and therefore no segregation for flower form can occur. 

 The reciprocal hybrid Wendlandii X Rexii has no anthers (see about 

 intersexuality in these crosses later) but can be crossed as a female 

 with Rexii X Wendlandii, which, as far as the genome is concerned, 

 amounts to the same F2 as before. In this F2 (also similar RF2) a 

 new recessive form of laciniated flower segregates ( choripetaly ) . 

 Obviously, the recessive mutant for this form of flower is present 

 in Rexii, but can become visible only in the presence of the 

 Wendlandii plasma, not in its own. The situation is similar to that 



