IMPROVEMENT OF OATS. 107 



selection of corn and cotton, and the same general forms of blanks are used for 

 making records. Numerous varieties were grown, and those that came nearest 

 to the different ideals were saved for selection, while those that seemed to be 

 inferior were discarded. 



As yet the work is confined to the Arlington Experimental Farm, near 

 Washington, but later on will be carried to other oat regions. One important 

 problem taken up is the selection of a good winter oat, the qualities to be 

 selected for being increased resistance to cold winters, rust resistance, in- 

 creased yield, strength of straw, size of grain, reduced percentage of hull and 

 the percentage of empty hulls, and the absence of awns. This will be done 

 by means of the score card system used with cotton and corn. The best 

 yielding plants, with a small percentage of rust and showing a tendency not 

 to lodge, were selected from the nursery and kept for further planting and 

 further selection the next year. Acre plats of a number of winter varieties have 

 been sown for selection next season, and as the Experimental Farm is just 

 on the border of the winter oat region, the killing out of the weak plants by 

 cold will aid materially in the selecting of hardy plants. The individual plant 

 is taken as the basis of selection, and large plats of single grain hills are 

 sown for this work. The plats of the past season were planted with a six-inch 

 space each way, but this fall they are being planted six inches apart in rows 

 with an interval of eighteen inches between rows, affording an opportunity for 

 close observation and also for cultivation. The individual plant system is 

 absolutely necessary for good work in selection on account of the difference in 

 stooling, evenness of ripening, etc. In the Virginia Gray winter plats a differ- 

 ence of three weeks was noticed in the ripening period of individuals planted 

 singly, while in drilled plats it was impossible to judge earliness except by the 

 ripening of single heads. 



One of the greatest annoyances is the difficulty of procuring pure seed. 

 Among all the samples planted last Spring but one or two were what could 

 be called pure. Most of the varieties are made up of two or more strains 

 varying in character. Often there are what appear to be mixtures of almost 

 all the types of oats in one lot, sometimes to such an extent that it is difficult 

 to tell to what the name is to be applied. Under these conditions it has been 

 necessary to go over the lots and carefully sort out the seed for next season. 

 All of these selections w-ill be planted next spring, and their progeny will be 

 selected, and in this manner pure types will be maintained for comparison 

 and to obtain known parents for crossing. 



On account of the complaint of many stablemen that most of the oats on 

 the market contain a large percentage of empty hulls, a selection has been 

 started for a type in which all of the flowers set seed even under adverse 

 circumstances. The rust-proof varieties seem to offer the best type for this 

 selection under the conditions that prevail on the Arlington Farm. 



Another selection started is that of reselecting some of the new varieties 

 originated in Europe to better fit the new conditions in this country. These 

 varieties, so far as tested, seem to sport, or break up into different types. 

 Carton's Tartar King was planted last spring and studied in order to find out 

 if possible whether it was thoroughly fixed in type. The plants were prac- 

 tically uniform in leaf and stem characters, but varied in some cases, quite 



