C. A. BARBER 8l 



immediately preceding them. The change is rather a sudden one. We shall 

 see, in our dissections, that, in any system of branching, while the as or main 

 shoots do not differ very greatly from the hs or branches of the first order, 

 there is usually a more marked difference between these two and the rs, the 

 latter being generally considerably thicker. It appears to us that, with the 

 favourable season of 1894, it is probable that in some of the 20 plants a branch 

 of the second order may have matured ; while it is highly improbable that any 

 of the 1895 shoots of the same order would be sufficiently advanced to be cut 

 as canes at crop time. In this case we should have, among the 1894 plants, 

 at least two which had a c branch matured, as is suggested in Fig. 3 of the 

 diagram, and this would explain the presence of these abnormally heavy canes 

 at the end of the season {cf. PI. XXXII, where diagrams of thick cane branch- 

 ing are shown, and Table on p. 147, where the formulae of Louisiana Stri2)ed 

 a,\\d Louisiana Purple, grown in the Cane-breeding Station, are given). 



We think that Stubbs is justified, from his figures, in assuming that, in the 

 Louisiana climate, the mother shoots are at harvest richer than the branches, 

 but we do not think that his facts are sufficiently convincing to assume 

 the regular decrease in sucrose among the younger shoots, excepting where they 

 are also increasingly immature, as in the 1895 crop. We also do not agree with 

 him in his contention that his figures give ground for the assumption that there 

 is a gradual decrease in the weight of branches in the order of development, 

 as it runs counter to all our experience in the dissections which we have carried 

 out, and is not borne out by the figures in his tables. Stubbs's paper is of 

 special interest to us, in that it is the only one we have met with in which any 

 care has been taken to separate the branches of different orders. 



The next pieces of work on the effect of spacing on the number of canes 

 produced are in 1910, when independent experiments were conducted by 

 Kilian and MuUer von Czernicki in Java. Kilian's experiments^ were made, 

 with J. 247, a late but good tillering variety, on dry loam, " strugge^ " loam 

 and heavy black clay. It is unfortunate that the control plot of the latter 

 was destroyed by fire ; this class of soil, namely heavy clay, is apparently less 

 suited to J. 247, and the results recorded of the single experiment show that 

 some unmentioned factor has intervened. This plot we have accordingly 

 left out in the discussion, and confined our attention to the four others, on loam 

 of varying fertility. Kilian planted his sets in rows 3J', 4' and 5' apart, and a 

 summary of his results is given in the appended table, averaging the duplicate 

 plots. 



1 Geerligs, H. C. Prinscn. The Worlcr^ Cane Sugar IiuJiislnj, Pasl ami rrea., p. 193, I9l2. 



2 Wo have been unable to translate this wurd, but imagine that this ioim is less lyitile. 



