VARIATION IX THE CAXE AND CANE VARIETIES 43 



Tanna, as sporting from a White Tanna. In addition, during the writer's 

 residence in Mauritius he was shown a number of "varieties" — possibly 

 identical and certainly closely allied to which the name Louzier rayee was 

 applied. In such a case a cane variety ma\' be regarded as throwing a 

 limited number of distinct sports, or the observation may be indicative of a 

 mixed cultivation of barely distinguishable varieties. 



Asexual variation or sporting is recognised only when some prominent 

 characteristic, such as colour, varies. Sporting however may very well occur 

 in the absence of means of read}' identification, and possibly valuable strains 

 or varieties have thus arisen, and continue to arise, but have been lost for 

 lack of means of recognition. Correlated with the colour change, other 

 characters also vary. Thus the \Miite Tanna has become cultivated as a 

 valuable cane, the Black Tanna being rareh' found in extensive cultivation. 

 Local conditions also seem to determine the economic value of the twin 

 sports; thus the dark-coloured sport from the Cheribon, Java, etc., cane has 

 been established in Java and Louisiana whUe in Cuba and the West Indies 

 generally it is the light-coloured sport which is favoured. 



A sporting habit would account, too, for the minor differences, such as 

 absence of setae on the leaf base, to be found on individual canes in fields of 

 supposedly pure cultivation. Further, as already micntioned, sexual descend- 

 ants of a variety generally resemble the parent or parents ; accordingly, 

 as has been suggested by Harrison^", a plantation grown from a pure stock 

 may in time become made up of the original stock, and of those sexual 

 descendants which resemble the parent. Such a state of affairs, due to the 

 combined influence of sporting and the presence of seedlings, may account 

 for the different results obtained in different districts from what is thought to 

 be one and the same variety. 



In the instances discussed at length above, and which have afforded 

 very valuable varieties of cane, a distinct and easily recognised variation 

 serves to establish and to fix the variety. Variation, however, may and 

 probably does occur without any easily recognized sign whereby it may be 

 identified. Attempts, almost entirely confined to Java, have been made to 

 correlate certain features of the cane with valuable characteristics, and in 

 this way to obtain new varieties or rather strains. These attempts include 

 the following features : — 



Disease immunity and inheritance. — As regards inheritance of sereh and 

 yeUow-stripe disease it has been found that the incidence of the disease 

 tends to decrease when disease- free cuttings are used for seed, the reverse 

 action obtaining when unselected tops or tops from diseased stalks are used. 

 This observation has been of great value in Java. 



Sugar content. — Based on the knowledge that the seed from sugar-rich 

 beets afford a rich strain, attempts have been made to obtain sweet strains 

 of cane by selecting for use as tops cuttings from sweet canes. Early re- 

 sults in the West Indies gave no promise of success, but Kobus^^ in Java 

 obtained in experiments definite results, and further observed that the 

 heaviest canes were the sweetest, so that the routine of the selection was much 

 simplified. Following on his work, Nash^^ and others in Java selected tops 

 on a specific-gravity basis, beheving that the descendants of such tops 

 would maintain that characteristic combined \v'ith a high sugar content. 

 The whole question has been the subject of further detailed studies, and of 

 much controversy in Java, %vith the unhappy finding that this means of 



