84 PHYSIOLOGICAL RELATIONS 



In all cases inoculation experiments should be made in quantity, 

 and control experiments in similar number must be relied upon to 

 eliminate any possibility of error. If a given disease is particu- 

 larly abundant in the region, and accidental infection therefore 

 more probable, the number of control cultures should be further 

 increased, in addition to the special precautions mentioned. 



A failure to secure infection from a relatively small number of 

 experiments may not indicate that the particular fungus plays no 

 part in the production of the disease with which it has been asso- 

 ciated. At any rate, experience in pathological work is necessary 

 when one assumes to make a positive statement in this regard. In 

 some cases infection may occur at a definite period only, or closely 

 related species of fungi may differ markedly with respect to the 

 conditions under which infection may take place. It has been 

 found that the fungus causing fruit spot of apple is effective at 

 about the time that the hairs covering the surface of the young 

 apple are broken off. The loose smut of oats penetrates the host 

 only when the latter is in the seedling stage, while the smut of 

 wheat may infect the blossom. 



