T. M. SONNEBORN 167 



with X, and moderately with XVI. The uniquely defined system 

 of reactions of each mating type, including of course the total 

 absence of reactions between mating types of certain different 

 varieties, was the original, and remains the most reliable, basis 

 for distinguishing the varieties. 



Isolation of the Varieties. The preceding account has passed 

 lightly over the important fact that conjugation can occur be- 

 tween certain different varieties. Since varieties 1, 3, 5, and 7 

 can interbreed, do they not have a common gene pool? Should 

 they not all be grouped together as one biological species? More- 

 over, since variety 4 conjugates with variety 8 and the latter con- 

 jugates with variety 3, should not varieties 4 and 8 also be mem- 

 bers of this species? These questions place squarely before us the 

 problem of what the common gene pools in P. aurelia really are. 

 To this problem, which includes the questions just asked, but is 

 broader than they, we now turn. 



First of all, there seems clearly to be no gene flow from cer- 

 tain varieties to certain other varieties. The seven varieties not 

 listed in Table I have never been observed to give the slightest 

 sexual reaction to animals of any variety other than their own. 

 This may be due to inadequate study so far as the varieties which 

 are newly discovered are concerned. However, others, like varie- 

 ties 2 and 6, have been long known, and it is practically certain 

 that neither of these reacts with the other or with any of the 

 other long known first eight varieties; gene flow between each 

 of these and the other six must be considered as nonexistent. 



Secondly, even when the bar to interbreeding is let down, as 

 among most of the varieties listed in Table I, other mechanisms 

 interfere with gene flow among them. The only open question is 

 whether the interference is completely effective or whether it is 

 not quite completely effective. Crosses between varieties 4 and 

 8 almost always end in death after a few fissions ( Sonneborn and 

 Dippell, 1946; Melvin, 1949). Levine (1953) thoroughly studied 

 the small proportion of Fl survivors. Although he could not prove 

 their hybrid nature, about a dozen among over 300 Fl clones 

 examined could have been true hybrids. Later generations, even 

 by backcrosses, from these rare survivors were, however, as 



