LECTURES IN BIOLOGICAL SCIENCES 



direction for many generations, even if no new mutations oc- 

 cur at all. 



Recent selection experiments with corn, Drosophila, mice, 

 chickens, and other cross-fertilizing populations of higher plants 

 have consistently shown that such populations are capable of 

 responding to selection for many generations on the basis of 

 the genetic variability stored in them. Two good examples are 

 the Illinois corn experiment and the selection experiments by 

 Mather and his associates for the number of abdominal bristles 

 in Drosophila. 



The Illinois experiment, described by Woodworth and Jugen- 

 heimer (1948), was as follows: In 1895 the agronomists at the 

 University of Illinois decided to find out the number of genera- 

 tions in which they could produce a change in a characteristic 

 by continuous artificial selection, using a cross-fertilizing popu- 

 lation of field corn. They selected for four different character- 

 istics—high protein of the kernels, low protein, high oil content 

 of the kernels, and low oil. This experiment was continued 

 for fifty generations; the most recently published results were 

 obtained in 1959 (the experiment was briefly interrupted during 

 the last war). In all four lines the population responded to se- 

 lection for at least thirty-five generations. In the case of high 

 oil and low protein, a significant change took place in the popu- 

 lations even between the forty-fifth and fiftieth generations. 

 During the experiment, the protein content was more than 

 doubled in the high-protein line and reduced to less than half 

 of the original concentration in the low-protein line. Even 

 greater results were obtained by selecting for high and low oil 

 content. The kernels of the original population contained 4.7 

 per cent of oil. After fifty generations of selection, the mean oil 

 content in the high line was raised to 15.4 per cent, and in the 

 low line it was lowered to 1.0 per cent. 



The facts which we know about mutation rates in corn tell 

 us definitely that the Illinois agronomists must have been sort- 

 ing out genetic differences which existed in their original popu- 

 lation. Since in each line of corn being selected the number 

 of plants raised per generation was between 200 and 300, the 

 total number of plants raised during fifty generations in each 

 line was between 10,000 and 15,000. We do not know the 



44 



