480 



SCIENCE 



[N. S. Vol. XSXIX. No. 1005 



of a particular organism in large numbers, 

 and by continued selection of the varia- 

 tions best meeting the conditions move 

 rapidly forward by a series of increments 

 toward the goal of perfection. What could 

 be more simple ? Instead of corn having an 

 acreage yield of fifty bushels, there would 

 with a proper supply of plant food be a 

 production of two hundred, two hundred 

 and fifty or even three hundred bushels. 

 Instead of politicians with no perspective 

 beyond their immediate welfare — a reelec- 

 tion — instead of college presidents and 

 faculties with their numerous shortcomings 

 — according to the students and occasion- 

 ally the trustees — there would be the ideal 

 individual bred to specification and not 

 necessarily made in Germany. 



Unfortunately, variations with a per- 

 verseness incomprehensible uniformly re- 

 fused to accumulate in the manner desired 

 and at times even demonstrated their obsti- 

 nacy by retrogression. It was plainly evi- 

 dent that there were limits imposed by 

 nature not easily passed, and in connection 

 with which much experimental work must 

 be undertaken before definite progress was 

 made and the facts fully understood. 



With a realization of the difficulties in- 

 volved in an attempt to apply evolution, it 

 will be well to pause for a moment and 

 consider certain fundamental principles 

 before discussing the results of some of the 

 investigations which for a time at least 

 promised much toward the solution of the 

 problem. Thus it may be stated that evolu- 

 tion in its different modifications postulates 

 in general (1) the occurrence of numerous 

 varying individuals, some of which are (2) 

 eliminated by environmental stimuli leav- 

 ing few or no offspring, while (3) the sur- 

 vivors transmit to their progeny the char- 

 acters which proved of selective value, with 

 the result that (4) through the continuation 

 of the process the race eventually becomes 



adapted to surrounding conditions. The 

 first two propositions are merely state- 

 ments of fact. The real difficulties of the 

 situation are those of ascertaining how 

 variations which are transmitted may be 

 ■recognized and produced so that the result 

 will be a cumulative one. Until this is done 

 breeders must continue to proceed in the 

 same haphazard manner that they have fol- 

 lowed for countless generations. 



By selecting the largest and most perfect 

 ears of seed corn from the variations pres- 

 ent in the field, conversely eliminating the 

 remainder from reproducing, the corn 

 grower plants with a fatuous trust in provi- 

 dence that a crop somewhat better or at 

 least as good as the preceding crop will be 

 produced. If it is a type comparatively 

 pure the average may be maintained and 

 the hope partially realized, but the chances 

 for retrogression are far greater than for 

 advancement, inasmuch as there is no means 

 for distinguishing a variation which will 

 be transmitted with equal or better results 

 than in the preceding generation, from 

 one that represents a fluctuation due to 

 nurture and which is non-transmissible. 

 Thus the apparently inferior ear of 

 corn will frequently produce a yield 

 far better than obtained from one which 

 is perfection as graded by the meth- 

 ods of the "corn show," and if from the 

 same pure race, the resultant crop wiU be 

 at least as good. Artificial methods of 

 hybridization, which furnish an immediate 

 advancement in the succeeding generation, 

 result in a gain which is only temporary. 

 The increased stimulus to growth vanishes 

 as a fluctuation. 



Thus it is quite evident that there exists 

 a problem in the evolutionary control of 

 organisms even the partial solution of 

 which will mark an extraordinary advance- 

 ment not only for agriculture, horticulture, 



