100 ^lASSACHTSETTS HORTICULTURAL SOCIETY. 



In strain breeding, on the contrary, while the method of selection 

 may be the same, the characters selected are the common characters 

 of the race and no new permanently heritable character is added. 

 The selection of the sugar beet for increased sugar content is an 

 illustration of this tj'pe of selection. As above described, in the 

 sugar beet, for many years selections have been made of mothers or 

 seed producers richest in sugar content and while this selection has 

 been under way for over sixty years the character of increased 

 richness of sugar content has not become fixed as a permanent 

 character of the race, but each year the selection of the mothers 

 richest in sugar content must be made if the industry is to prosper. 

 Here we would appear to be keeping up a strain Avithin a race by 

 the continuous selection of the maximums. 



In the ordinary work on corn breeding which has been conducted 

 to secure primarily, increased yield without reference to great fixity 

 of type, it is probable that the results are mainly the same as in the 

 sugar beet and that the selection must be continued regularly to 

 retain the desired improvement. The methods pursued by stock 

 breeders are almost entirely what we have called strain breeding 

 methods. The stock breeder rarely, almost never, attempts to 

 produce new breeds. He selects within the breed to maintain the 

 highest strain of the breed. The good stock breeder well knows 

 that he must select each generation carefully or rapid deterioration 

 or regression toward the mean of the race results. 



If the plant breeder is seeking to produce new races, such as the 

 Columbia cotton, he will search his fields of whatever plant he has 

 under consideration for variations indicating new characters of 

 value. These variations he will isolate to test and select to fix the 

 characters. He will, in any case, naturally select the best individ- 

 uals each generation for seed, taking them from the best progenies. 

 If there is any cumulative effect of the selection, so much the better. 

 If there is no cumulative effect he will at least have purified his type 

 by his selection and, meanwhile, will have determined whether his 

 new type is any improvement on the old forms and whetlier it is 

 worthy of preservation. In either case the selection is practically 

 indispensable and he must finally prove his new sort either worthy 

 or unworthy, and on his judgment of this matter his skill and safety 

 as a breeder rest. 



