J. A. MOORE 329 



so well with geographic distribution that it seemed clear there 

 must be a causal connection. The exceptional species was R. 

 pipiens, which had a geographic distribution much greater than 

 could be anticipated on the basis of the temperature adaptations 

 of individuals from the New England area. 



Studies of R. pipiens from south of their predicted geographic 

 boundary showed that these southern forms differed from their 

 northern counterparts in their temperature adaptations. This 

 situation is by no means surprising. One would anticipate that 

 a widely distributed species would be broken into local popula- 

 tions, each adapted to its immediate environment. R. pipiens 

 embryos adapted for early spring breeding in the cold-temperate 

 environment of Quebec and northern New England would not be 

 expected to survive under the temperature conditions encoun- 

 tered in the semitropical zones of southern Florida or eastern 

 Mexico. 



For our present purposes it is of considerable interest to note 

 that developmental (and presumably genetic) incompatibility 

 is associated with adaptation to different temperatures. If in- 

 dividuals from similar temperature-adapted populations are 

 crossed, the embryos are normal in their development. On the 

 other hand, if individuals from different temperature-adapted 

 populations are crossed, the embryos are very abnormal and 

 may die early in development. 



The extreme results are obtained in crosses of Vermont x 

 Florida, Wisconsin x Texas, and Vermont x eastern Mexico in- 

 dividuals. In the last cross most of the embryos die as abnormal 

 gastrulae or neurulae. Thus, these southern pipiens and northern 

 pipiens are behaving as different species so far as their potential 

 ability to exchange genes is concerned. These are phenomena 

 associated with the crossing of individuals from widely separated 

 latitudes. Adjacent populations, however, can be crossed, and 

 they give entirely normal offspring. 



We could consider the populations of R. pipiens to be one 

 species, two species, or to be one species in a crucial stage in the 

 process of splitting into different species. I believe the third 

 alternative is the most meaningful way to consider the data. The 



