[Vol. 2 



278 ANNALS OF THE MISSOURI BOTANICAL GARDEN 



of no avail. Losses of 25 per cent of the crop are not uncom- 

 mon. The solution of this problem seems possible only by 

 the breeding of disease-resistant varieties. It is certain that 

 smut-resistant races of wheat exist. The problem is to find 

 these varieties and, in case they are not sufficiently produc- 

 tive, to cross them with other varieties until races which com- 

 bine the desired characteristics are obtained. In the districts 

 where smut occurs every year it is possible to find these races 

 in an empirical way. But in general it is my opinion that all 

 work of selecting and breeding should be prosecuted along 

 fundamental scientific lines. 



It is therefore first of all necessary to determine to what 

 characters the plant owes its disease-resistant qualities. 

 When this has been accomplished it is next necessary to deter- 

 mine to what extent the characters are heritable, that is to 

 say, whether they appear in crosses as dominant or recessive. 

 The great advantage of this method lies in the fact that it 

 makes it possible to recognize resistant races (by the presence 

 of the specific characters to which resistance is due) without 

 infection experiments, which are uncertain owing to the influ- 

 ence of external and unknown conditions. 



I have shown to you by this example that in the solution 

 of a single phytopathological problem such diverse branches 

 of botany as taxonomy, biology of the flower, fungus-biology, 

 and inheritance are involved. The following examples will 

 show that in addition other branches of botany are of import- 

 ance in phytopathology. 



In exact phytopathological investigations it is a primary 

 factor that one know the host plants and the parasites in 

 detail. This information must be based upon thorough 

 systematic knowledge. This seems to be very easy in culti- 

 vated plants, the species of which are generally well distin- 

 guished. Some cases, however, are more complicated. When 

 we want to make studies of cereal rusts, it is not sufficient 

 to know the races of cereals by their agricultural names. We 

 must know to what botanical species they belong; our culti- 

 vated wheats, for instance, comprise species of different sus- 

 ceptibilities. 



