The Cytoplasm as Specific Substrate 225 



term 'loss mutation," and so on, the danger increases that superficial 

 readers of this beautiful work are swept oflE their feet by the termi- 

 nology and believe that here "plasmagenes" and their mutation have 

 been demonstrated. I think that a careful consideration of the aphid 

 example just discussed will show that the facts are rather simple and 

 that there is no need to complicate the explanation by introducing the 

 wobbly concept of plasmagenes, as some have done in referring to 

 these facts. We should think twice before describing as mutation 

 special features of cytoplasmic division involving unequal distribu- 

 tion of cytoplasmic components, and thus intimating that we are 

 dealing with a process comparable to mutation of genie material. 

 Actually, the fact of a 100 per cent mutation rate should suggest 

 caution. This conclusion is independent of the details of the action of 

 the active material (in this case the euflavine) upon the cytoplasmic 

 constituents, whether it affects only their quantitative distribution, or 

 sorts out different kinds of them, or dissolves, poisons, and aggluti- 

 nates them. (See Ephrussi's detailed discussions.) It should again be 

 emphasized that here, just as in the former cases, the cytoplasmic ele- 

 ments under suspicion are the mitochondria, and "cytoplasmic action 

 on heredity" amounts to a generalized interference with the respira- 

 tory enzyme system (by removing the enzyme carriers) with conse- 

 quences for all genetically controlled processes which are sensitive to 

 such changes of what we may call, in a general way, the substratum. 

 (See discussion of Michaelis' work, II 2 C b.) 



An illustration of the good reasons for my warning against 

 wrongly suggestive terminology occurs in further work of Ephrussi 

 (see 1953). He found in a different strain what turned out to be a 

 Mendelizing mutant for respiratory deficiency. This could be crossed 

 to a clone, with that deficiency based upon the absence of the cyto- 

 plasmic carriers of the enzymes. The result was as expected: normal 

 Fi and 2:2 segregation in F2, since this was an ordinary MendeHan 

 cross and the zygote had the cytoplasmic particles from the parent 

 normal in this respect. But Ephrussi calls both strains mutants, and 

 therefore needs long explanations and many illustrations to explain 

 the simple fact, which is made complicated by the terminology he 

 uses. Of course the existence of these genuine recessive mutants shows 

 that the cytoplasmic particulates (mitochondria?) are not completely 

 autonomous, since a recessive mutant locus can prevent their function. 

 In principle this is not different from the fact that a recessive mutant 

 in Drosophila prevents the formation of kynurenine from tryptophane. 

 The only difference, I think, is that in the first example the site of an 



