444 INTRODUCTION TO EVOLUTION 



and one blue bead) or nun (two blue beads). In "matings" like the second 

 listed, in which both parents were homozygous, the offspring must neces- 

 sarily both be of one constitution {Mm, in this instance); hence, no draw- 

 ing of beads was needed. 



In an actual experiment the offspring derived from the "matings" listed 

 above gave the following totals: \MM, lOMm, \mm (Fig. 19.1). 



The process was then repeated to produce a second generation. One 

 pair of red beads, ten pairs consisting of one red bead and one blue bead, 

 and one pair of blue beads were put in the box, mixed, and withdrawn 

 at random, two pairs at a time. Thus "matings" were arranged by chance. 

 Then the two offspring from each "mating" were determined, using the 

 bowl of red and blue beads as described above. 



The process was repeated "generation" after "generation." Sometimes 

 the number of blue beads increased, sometimes the number of red beads 

 did so. Sometimes beads of one color almost disappeared, the population 

 coming to consist of 1 \MM and \Mm individuals. Then there would be a 

 "rally" on the part of the m genes, which became more numerous again. 

 The model demonstrated the tenacity with which the laws of probability 

 expressed in the Hardy-Weinberg formula tend to maintain equilibrium 

 even in such a tiny "population." Yet eventually "drift" won out, and one 

 gene was entirely lost. The first time the experiment was tried, fixation of 

 one gene did not occur until the 134th generation; members of that genera- 

 tion were all MM (12 pairs of red beads ) . That the large number of gen- 

 erations was of no real significance, however, was attested by another run- 

 ning of the experiment in which the same result was achieved in seventeen 

 generations. These results are given in the accompanying table, study of 

 which will make clear that complete elimination of the blue beads might 

 easily have occurred in even fewer generations. 



What has genetic drift to do with evolution? As noted previously (p. 

 350), it affords a means by which inherited characteristics can become es- 

 tabhshed in a population without regard to their usefulness. When the size 

 of the population is small some genes may be lost or reduced in fre- 

 quency by chance, others may be increased in frequency by chance. Thus 

 the nature of the population is changed without involving the matter of 

 usefulness. In our discussion of classification (Chap. 14) we noted that 

 the structural differences between species, and especially those between 

 subspecies, are frequently small. Commonly, also, these differences do not 

 seem to be of importance to their possessors. What difference does it 

 make to a mouse whether the margins of its dorsal tail stripe are clearly 

 drawn or indistinct (see p. 316)? Genetic drift affords a means by which 



