THE ORIGIN OF VARIATION 



OLYMPIC <A) 



OAXACA(A) 



HIDALGO (A)« 



MAMMOTH (A), 



•TREE LINE (A) 



V 



-> ESTES PARK (A1 



CHIRICAHUAKA) 



SANTA CRUZCA) >■ CUERNAVACA (A) 



HYPOTHETICAL- MIRANDA 



TEXAS fA)-<— PIKES PEAK (A)-< STANDARD (A&B) > ARROWHEAD (Al >CH1RICAHUAII (A) 



SEQUOIAKB) 



KLAMATH (B) MENDOCINO (B) 



COWlCHAN(B) WAWONACB) SEQUOIA E (B) HUMBOLDT (B) 



Figure 85. Phylogenetic Chart of the DrosopliiJa pseudoobscum Group as Worked 

 Out by Analysis of Overlapping Inversions. Race A of this diagram is D. pseudo- 

 obscura and race B is D. persimilis of the newer classification. (From Dobzhansky, 

 "Genetics and the Origin of Species," Columbia University Press, 1937. ) 



THE ARCHITECTURE OF THE CHROMOSOMES 

 AND SYSTEMIC MUTATION 



Goldschmidt urged that only two possibilities exist with respect to the 

 significance of such chromosomal diflFerences between closely related spe- 

 cies. Either they are simply one among the many phenotypic characters 

 which differentiate species, having no more causal significance than such 

 characters as coat color, tooth specialization, or form of ovipositors; or else 

 they constitute the actual genetic basis for the separation of species. He 

 favored the latter viewpoint. 



Goldschmidt's reasons cannot be treated in detail here, but some indi- 

 cations of the basis of his viewpoint must be given. At the outset, he 

 rejected the corpuscular theory of the gene, which is basic to the neo- 

 Darwinian theory. He did this because he believed that the principle of 

 parsimony demands that the important facts of both gene mutation and 

 position effect must have a single explanation. Very briefly, because gene 

 mutations and chromosomal rearrangements respond to radiation simi- 

 larly, because position effects resemble the effects of a gene near the break 

 and may behave as an allele of it, and because all types of genie action 

 also occur as position effects, Goldschmidt believed that these two phe- 

 nomena were one. Further, because a position effect sometimes overlaps 

 two or more genes, because breaks anywhere in a small region of the 

 chromosome may give the same position effect, and because position ef- 



224 



