DE VRIES'S THEORY OF MUTATIONS. 209 



tinguished : one, which is called fluctuating variability, oscillates round 

 a mean value; we shall consider it a little more closely. Whichever 

 characteristic of a species we happen to choose, we shall always find, 

 in considering a number of specimens, that individual differences, indi- 

 vidual variations, can be noticed which, when tabulated according to 

 size or to number, do not exceed two opposite extremes. Half-way be- 

 tween these extremes we find the 'norm' for that particular character- 

 istic. The fluctuation may be represented by a curve, the culminating 

 point of which corresponds to the norms just mentioned, whereas to the 

 right and to the left of it the curve gradually approaches the hori- 

 zontal line and has a symmetrical shape. Quetelet and Galton have 

 insisted on the great significance of the fact, that fluctuating variation 

 remains enclosed within the limits of such a curve of regular shape; 

 the curve itself is, therefore, often called Galton 's curve. 



Not only for plants, but also for animals and especially for man, 

 the existence of such Galton curves, expressing the amount of varia- 

 bility, has been definitely established by different observers in very 

 numerous instances. Thus, for example, Ammon has obtained his 

 material from South German recruits. We need not insist on the fact 

 that the greater the number of cases, the more reliable the curve. 



The different degrees of fluctuating variability can undoubtedly be 

 seized upon by any one who wishes to make them the starting-point 

 for the breeding of certain distinct variations. Thus, for instance, by 

 constantly selecting for the reproductive process those plants in which 

 a given deviation is strongly marked, after a certain time and after a 

 series of generations, a plant can be obtained for which the Galton curve 

 would indicate a displacement of its culminating point in the direc- 

 tion of the selected variation. In this way an increase in the yield of 

 sugar obtained from the beet roots has been arrived at from about 7 

 per cent, to 13 or 14 per cent. Thus also ears of maize have been pro- 

 duced that bore 20 rows of grain, whereas the kind from which the 

 experiment had started always bore 12 to 14 rows. 



As soon, however, as such conscious and voluntary selection ceases, 

 the next generations successively return to the original curve. In 

 order to prevent this retrograde process, without a constant and re- 

 peated application of the artificial selective process, we are obliged to 

 prevent the appearance of new generations, by forcing the plant to 

 reproduce itself not by seed, but asexually by means of buds. It is 

 well known that definite kinds of delicate fruit are reproduced in this 

 fashion, because if multiplied by seed, they would always tend to fall 

 back into their former state of less value. Transposing the culmina- 

 tion of the curve of variability artificially, as explained, or breeding 

 variations to the right or to the left of the norm, can never exceed 

 certain limits. Agencies are at work there which prevent the fluc- 



VOL. LXV. — 14. 



