SUMMAEY. 77 



the 3-month and 7-nionth periods preceding narvest, it is seen that 

 1905 had more precipitation than 1903 or 1904; the relative humicHty 

 was greater in 1905, but the average temperature, though about the 

 same for the 7-month and S-month periods chu-ing the 3 years, 

 averaged 2.67 degrees subnormal over the whole area in 1904 

 during the month containing the critical period. It was 3^ degrees 

 below normal in South Dakota, North Dakota, Minnesota, and Wis- 

 consin, the region most affected by rust in 1904. This average was 

 considerably lower than that of the same period in 1903 and 1905 

 over the same area. It is believed that this unusually low tempera- 

 ture in 1904 was a very important factor, if not the determining 

 factor, for the rust epidemic of that year. 



(8) Spraying experiments for the prevention of rusts have been 

 tried from time to time by various investigators, but for the most 

 part without satisfactory results. There is doubt that spraying will 

 ever be of practical value for rust prevention, but as the critical 

 period for wheat, with regard to the attack of stem rust, is now 

 known, further spraying experiments limited to this period may give 

 valuable results. 



That excess of some elements in the plant food may predispose a 

 plant to disease or render it more resistant has not been firmly estab- 

 lished. That indirectly it will have some influence, by affecting either 

 the physiology or the general growth of the host plant, is very 

 probable. 



Wliere soils are rich in nitrogen, other conditions being equal, rust 

 attacks are, as a rule, most prevalent. 



Experiments in soil treatments for disease prevention have been 

 made by various investigators, but no very practical results have 

 been reported. This field of work is promising and should be further 

 investigated. 



Some plants are more resistant to attacks of parasitic fungi than 

 others, and it has not yet been definitely established to what character 

 in the plant this resistance is due; but most authorities agree that 

 resistance is due, as a rule, not to morphological but to physiological 

 characteristics. 



Disease resistance is inheritable to a greater or less degree, and 

 Biffen has brought forth experimental results to show that resistance 

 and susceptibility of cereals to rust are Mendelian characters. 

 Other investigators have reached similar conclusions. 



There are three methods in use for developing rust-resistant grains 

 through selection and breeding: (1) Testing and selection of pure 

 varieties to determine which are resistant; (2) selection of the best 

 individuals, or bulk selection from some strain or variety from year 

 to year under fairly constant disease conditions; (3) hybridizing of 



216 



