BIOLOGY AND HUMAN TRENDS — PEARL 331 



ment designed to test this point in a simple and direct way. To- 

 night I make the first public statement about it. This experiment 

 has now included over 300 successive generations — perhaps the long- 

 est bit of controlled breeding ever carried out, with the results in 

 each successive generation carefully observed and precisely recorded. 

 Allowing 30 years as a round figure for the average duration of a 

 human generation the time equivalent in human reproduction of this 

 experiment would be of the order of 9,000 years — considerably longer 

 than the total span of man's even dimly recorded history. The ob- 

 jective of this experiment with Drosophila has been to see whether 

 a simple Mendelian ratio involving but one character would or could 

 be altered in the passage of time by such natural forces as selection, 

 different sj^stems of breeding (such, for example, as that called 

 "grading up" by livestock breeders), and wide alterations of the 

 environment nearly up to the limits of the organism's ability to go on 

 living at all. The plan of the experiment is a simple one. It started 

 by crossing a normal fruit fly {DrosojJhila melanog aster) possessing 

 the normal wings characteristic of the species, with the pure mutant 

 form Vestigial^ so-called because the wings are reduced to nonfunc- 

 tional vestiges. This wing characteristic is associated with a single 

 gene. In the next generation all the flies produced by the pair with 

 which we started had normal standard wings, normal being domi- 

 nant to vestigial. These flies of the first cross-bred generation were 

 then mated to pure vestigials (back-crossed to the recessive parent, 

 in technical genetic language) to produce the second cross-bred 

 generation. Of the offspring of these matings approximately one- 

 half had normal wings, because they carried the original normal 

 wing gene, and the other half had vestigial wings, all this being in 

 accord with regular Mendelian expectation. The vestigial-winged 

 flies of this and all later generations were killed and thrown away 

 as soon as the}^ had emerged and been counted. The normal winged 

 flies were again mated to pure vestigials to produce the next genera- 

 tion. And so on with undeviating regularity for more than 300 

 generations. What the plan means in briefest terms is that since the 

 rather stupendously long time (measured in generations) when the 

 experiment began the only hereditary determiner (gene) for normal 

 wings that has ever been in the system is the one that was con- 

 tributed by the one single normal wild type fly with which we started. 

 All the normal winged flies now appearing in the populations of the 

 successive generations of the experiment have normal wings only 

 because their Urgrossvater had them 300 generations ago, and for 

 no other reason. 



The net result of the experiment has been to show that the gene 

 involved has preserved its initial characteristics unaltered. So also 



