The following paper was read by S. A. Beach: 



CORRELATION BETWEEN DIFFERENT PARTS OF THE 



PLANT IN FORM, COLOR, SIZE AND OTHER 



CHARACTERISTICS. 



S* A. Beach t Horticulturist/ New^ York State Experiment Station, Geneva, N*Y. 



Work in plant breeding, when considered from the standpoint of the 

 worker, falls into two general classes. In one class the effort is to originate 

 an improved strain or variety; 'in the other the object is to learn the phil- 

 osophy of plant breeding, to discover the scientific principles involved and to 

 illustrate the application of these principles. Immediate practical results are 

 sought in one case; in the other a knowledge of the laws of plant breeding, 

 by means of which continued progress may be made. The former, if success- 

 ful, gives only something transient. The improved variety or strain which is 

 produced will doubtless be superseded in time by something better, as the 

 evolution of cultivated plants progresses. Such is the verdict of horticultural 

 history. We look to the other kind of effort for the more permanent and 

 eventually more rapid progress in breeding plants. It is the purpose of 

 this essay to call attention to a question, the investigation of which may yield 

 results of general importance and permanent value, namely, the extent to 

 which correlation between different parts of the plant in form, color, size and 

 other characteristics may be regarded as a significant factor in plant breeding. 

 Nurserymen, seedsmen, fruit growers and gardeners know many instances 

 of such correlation, and often take practical advantage of it when making 

 selections for propagation, albeit perhaps unconsciously, and without for- 

 mulating in words a definite expression of their judgment on this point. 

 Scientific investigators have also occasionally recognized instances where cor- 

 relation is of significance as a means of selection in breeding or propagating 

 plants, but it appears that this subject has not been investigated as system- 

 atically and thoroughly as it should be. 



In plant breeding the chances for originating an improved variety are in 

 some degree proportionate to the number of seedlings produced. Other 

 things being equal, the more numerous the seedlings subject to selection, the 

 greater the probabilities of finding the improvement sought. But the more 

 seedlings one undertakes to grow the greater the necessity of getting rid of 

 the undesirable ones at as young a stage of growth as possible, and thus 

 avoid the labor and expense of growing a great number of useless plants; 

 therefore, a skillful breeder does not defer the process of selection till the 

 seedlings appear, but exercises rigid choice, perhaps with the seed which is 



