BREEDING TIMOTHY. 61 



notes, recoi'ds, and pedif^ree summaries can be worked out for the 

 mother plants and their centgeners of progeny in careful breeding of 

 corn in much the same manner as has been outlined in the breeding 

 of Avheat. The yield, the per cent of nitrogen, the size, form, and char- 

 acter of the stalk are individual characters which should be numerically 

 recorded in securing superior parent plants; and in testing tliese 

 mother plants by comparing large families of their progen}- the cent- 

 gener yield, j)ercentage of nitrogen, etc., should be recorded, averaged, 

 and compared. 



Farmers can and do materially improve many of their varieties of 

 corn by crossing and selecting, and in some cases they injure good 

 varieties by injudicious crossing. Once a superior variety is obtained, 

 it should not be allowed to be mixed with another or replaced by 

 another without the best of proof that the change will be for the 

 better. 



BREEDING TIMOTHY. 



In 1889 the writer gathered seeds from numerous timothy plants 

 found along the wayside and on farms in the neighborhood of the 

 Minnesota experiment station. A hundred or so seeds from each 

 mother plant were planted in a plot, one seed in a hill, the hills being 

 12 by 18 inches apart. When these plants were 2 years old, each 

 having stooled, making a bunch a foot, or more across, the best plants 

 were chosen from the best plots. Seeds from these best plants were 

 harvested and plots were similarly planted, and this process was 

 repeated for succeeding generations. About 50 plants of the third 

 generation were divided up into settings, which were transplanted 

 into plots a square yard in area. When these were 2 years old, seeds 

 were saved from the best plots, and this seed was sown to increase 

 the stock of seed from these varieties. The variations among the 

 plants was sufficient to warrant us in attempting to select some of the 

 stocks as mothers for the development of meadow varieties and others 

 for pasture varieties. The seed has now increased to sufficient quan- 

 tity for making field tests of the yield of dry matter and the yield 

 of nitrogen per acre in meadow or pasture plots. 



It so happened that some of the veiy best plants had a tendency to 

 long spikeU^ts, and several of the 14 stocks which are being increased 

 in stock-seed plots have barbed spik(;s, such as are shown in Plate VI, 

 fig. 2. The tliree spikes on the right represent the foundation stock 

 from which was developed by selection tlie new timotliy represented by 

 the three si)ikeson the left, which shows a tendency to branch by length- 

 ening some of its spikelets. A distinguishing mark like this would 

 have value in a new kind of timothy, since it would distinguish it 

 from common timothy, which has not as yet been broken up into suc- 

 cessful varieties. But in the end the historical method by numbers 

 used for names may be the most practical way of keeping track of the 



