172 RED ROT OF StFGARCANE. 



a cylinder aseptically with a small cork borev and inserting a small 

 quantity of a pure cultme of CoUetotriehani. After eight days they 

 were examined. The lowei- inoculations were found to have infec- 

 ted 2 + 2 + 1 internodes respectively, the upper 1+2 + 4. The ex- 

 periment was repeated with Samsara canes, 6 being inoculated in 

 the same way. After seven days one was examined and was found 

 to have 2 internodes infected at the base and 1 at the top. Two 

 days later another was examined, 5 and 2 internodes respectively 

 being found infected. Two days later the rest were examined. In 

 one the infected portions had united in the middle. In tlie other 

 three, 11 + 10 + 7 internodes were found infected at the base and 

 3 + 5 + 2 at the top. The experiment on page 167, where inocula- 

 tions with the leaf form of the fungus were made at the to]) and 

 bottom of growing canes, should also be compared. As the prac- 

 tical point at issue was to determine if any recommendations could 

 be made for planting one part of the cane rather than another, 

 where red rot is prevalent, the natural inversion that goes on after 

 cutting was not taken into account, since it must equally go on in 

 planted setts before germination. The experiments do not suggest 

 that tops, though richer in glucose, can be more rapidly invaded by 

 Colletotrichum than the rest of the cane and there appears to be no 

 objection to their use from this point of view. The second experi- 

 ment and that given on page 167, indeed suggest that the contrary 

 is the case. Tops are also less likely to contain the fungus, when it 

 has originated from below at a late stage in the giowth of the cane. 



Control of the disease. 

 The conti'ol of red lot was stated by th(> eai-li<M' iiixcst igators 

 to be hkely to be veiy dinicult. owing to its posilion in llic inlci'ior 

 of the cane, the frequent absence of definite sviuptonis bv which il 

 might be detected in the growing crop and the practical impossibility 

 of preventing wounds which would give an entry to the fungus. 

 But of recent years little has been heard of the disease in Java, 

 where it was first described, and it may be concluded that it has not 

 proved so serious an enemy as was once feared. 



