34 PLANT BREEDING. 



were growing in large field plots. There was occasionally a verj^ tall 

 plant, and by diligent search plants were found which had a very 

 much branched top, and wei-e appai ently heavy seed bearers. A tall 

 plant and a heavy seed-bearing plant were chosen from each variety. 

 The next year 100 seeds were planted from each of these 14 mother 

 plants. The plants from the tall mothers averaged several inches 

 taller than those from the heavy seed-bearing mothers, tliough the 

 latter had not been chosen for low stature, and the supposition is that 

 they were of average stature among the plants of the respective varie- 

 ties from which they were selected. Thus, the first year the selections 

 were made among very large numbers, and the second year the mother 

 plants were tested by comparing a large brood of the progeny of each. 



In the accompanying illustration (PI. II, fig. 1) are shown the 

 results of the experiments in the improvement of flax by selection : 



The bundles at the right in the lower row are several samples of 

 flax bred for branching tops with many seed bolls, no effort having 

 been made to change the height by selection. 



The six bundles at the left have the same parentage as the seven 

 in the same row, but were bred by selection for greater length of - 

 straw. All the bundles of this row are from plots planted at the rate 

 of over 2 bushels of seed per acre, as in growing crops of fine fiber. 



Seeds from the same stocks were planted thinly, less than 3 pecks 

 of seed per acre, as for growing crops of flaxseed, and bundles of the 

 resulting plants are shown in the upper row. 



Whether grown thickly or thinly, the newly bred qualities of tall 

 growth and fine stems showed in the plants grown for fiber, which 

 stood several inches taller than those grown for seed, which remained 

 the same height as the plants of the original parent varieties. The 

 proof is very strong that long, fine fiber can be grown in the dry cli- 

 mate of the Northwest. Flax bred to grow tall will make fiber in 

 Minnesota closely approaching in length, fineness, and strength that 

 grown in western Europe under moister conditions. 



The most important feature of this experiment to Minnesota is the 

 yield of tlie fiax bred to grow seed. The tests of yield for seed have 

 not yet been completed, but since such a profound change has been 

 made in the height of those grown for fiber it would be very strange 

 did not some of the strains shown on the right prove very heavy 

 yielders of seed. 



From some of the best of these mother plants the experiment station 

 now has seed, as yet onlj- in sufficient quantity for field tests, of seven 

 strains of tall-growing kinds of flax, and seven others Avhicli are 

 apparently superior in their yield of seed. If the stocks first planted 

 in the field-crop nursery (among which relativelj^ little variation was 

 found because of the small number) had not been discarded and had 

 the mother plants been chosen from among them instead of from the 

 fields, where there was a large number to choose from, it is probable 



