476 SCIENCE PROGRESS. 



Cheribori in Java, for instance, being assumed to be near 

 perfection.^ 



Attempts at creating a disease-proof variety of " soft " 

 cane are more to the point. This has engaged much 

 attention in Java, and a certain amount of success has 

 already been attained. The method suggested is the 

 following : Select plants which were living as long as 

 possible before the first general appearance of sereh. The 

 chances are that such plants have a more or less resistant 

 character. Form a new plantation from the best of these, 

 using only perfectly sound canes for planting. Constantly 

 weed out those attacked by sereh." 



1 6. The principal objection to this method is its slow- 

 ness. A more rapid, but less reliable, means of obtaining 

 a new variety is by bud-variation. It occasionally happens 

 that a plant produces a bud quite different from the rest 

 in foliage and other characters. If the shoot thus produced 

 can be used for propagation, a new variety will probably be 

 obtained, because its peculiarities will remain fixed as long 

 as the vegetative method is adhered to. If, however, the 

 variety thus obtained is allowed to seed, it at once loses its 

 assumed characters, and tends to revert to the parent type. 

 It is exceedingly rare that such a variation is detected in 

 the sugar-cane, and the probability of its occurring in an 

 entirely useless direction practically removes bud-variation 

 from available methods. 



Another possible way of obtaining new varieties 

 asexually is by what are known as graft- hybrids. This 

 subject appears, however, to be rather obscure, and the 

 cases in which success is claimed do not always bear very 

 close examination. It occasionally happens that in grafting 

 a shoot arises from the point of union of stock and scion, 

 presumably by fusion of the parent cambiums, and unites 

 the characters of the parents. It will be seen at once that 

 graft-hybrids must be of rare occurrence, when we consider 

 that the operation of grafting is principally performed when it 

 is desired to perpetuate the characters of the scion unaltered. 

 There seem to be authentic cases of varieties arising in this 

 ^ See Sugar-Cafie, 1894, xxvi., 73 and 271. -Went (4). 



