CALDECOTT AND NORTH: RADIO-SENSITIVITY OF SEEDS 401 



our differing views on the motivation for doing the research reported 

 and the nature of the genetic systems involved. 



1. To test the hypothesis, or "assumption" to use your word, that the 

 hetero/ygote is superior to the homozygote is not, in itself, bad. One 

 could have set up the opposite hypothesis which would have been sub- 

 ject to the same kind of criticism that you render but which, I sense, 

 you would condone. 



2. The fact that overdominance has not been "convincingly" demon- 

 strated is the crux of the whole issue. II it had been clearly demonstrated, 

 or proved not to exist, both in diploids and polyploids (and I empha- 

 size the latter), I would have no interest in conducting the studies out- 

 lined as they would contribute nothing new to our knowledge. In this 

 regard, in my opinion, it is dangerous to generalize from what are admit- 

 tedly inadequate studies with two species, if you accept the negative 

 data in Drosophila, both of which are diploids, and neither of which 

 is naturally inbred or exists in nature at the polyploid level. 



3. You make a distinction between genetic diversity and epistasis. 

 These two concepts need not be, indeed they cannot realistically be, 

 treated independently. Therefore, it seems inconsistent to me that, on the 

 one hand, you postulate that heterosis may result from genetic diversity 

 but, on the other hand, not from epistasis. If you recall, in the text it 

 is specifically stated that the procedures being used should be of sig- 

 nificance if heterosis is due either to the differences between genes 

 governing the same order of biochemical function or to epistasis. 



MacKey: Wheat is neither strictly autoploid nor alloploid. It varies with 

 the characteristic. My own investigations indicate this rather clearly. 

 Characteristics essential for the ability to complete life and to repro- 

 duce show a high degree of autoploidy, while other characteristics not 

 decisive to life or death but determining morphology, adaptation, etc., 

 may follow a more alloploid pattern. The reason for this difference is 

 that the former category will have a more conservative differentiation 

 on the diploid level, while the latter may evolve rather freely as evident 

 from the morphology of Triticum monococcum, Aegilops speltoides, 

 and A. squarrosa, the parental genomes of 6x wheat. 



First, when diploids are added to form polyploids, the essential genes 

 can mutate more readily without severe deleterious effects, but also 

 here, at least, one pair has to be kept intact. The others are, however, 

 now free to mutate in directions that may have been impossible on the 

 diploid level. The degree of diploidization revealed in my experiments 

 with dicoccum favors the idea that this process has selective advantage. 



