318 



"strength" of the latex) from different trees, and appeared to 

 be constant and characteristic for the individual tree. Some 

 trees yielded only 23 grms. of rubber per 100 c.c. of latex, while 

 at the other extreme trees were found yielding as much as 

 54-55 grms. of rubber per 100 c.c. of latex, the mean for the 

 245 trees examined in this connection being 36-58 grms. per 



100 c.c. 



The author concludes from his observations on older trees, that 

 as a tree grows older the rubber content of the latex yielded by 

 it increases 1-2 per cent, per annum. The results are set out in 

 tabular form, and while only relatively few trees show the lower 

 percentages of from 23-29 grms. and the higher from 44-55 grms, 

 the majority yield amounts varying from 30-43 grms. 



With regard to yield of rubber, it was found that individual 

 trees were comparatively constant and from observations extend- 

 ing over two years it is stated that a tree which was found to be 

 a high yielder at one time could be relied upon to give a high 

 yield at all times. The yield-results are summarised in a table 

 and by means of a curve, and embody the examination of 1011 

 seven-year old trees. 



The mean yield in grammes per day came to 7*12 grms., but 

 some few trees yielded more than 2T grms. a day, and for a 

 large number the yield was from 0-2 grms. Thus from 9*6 per 

 . of the total number of trees 28 per cent, of the total yield 

 was contributed, while 13-7 per cent. (0-2 grms. group) only 

 gave 2*9 per cent, of the yield and certainly did not repay the 

 cost of tapping. Four outstanding trees in the plantation gave 

 41-45, 41-56, 41*72 and 42-77 grms. per day. 



The great possibilities of seed selection in improving rubber 

 yield are very clearly indicated from these figures. 



A further important observation in connection with seed 



selection to which. Mr. Whitby draws attention is that the seeds 

 from any one tree are exactly similar in appearance as regards 

 tint, mottle pattern and shape. 



It now remains to be seen whether seeds from a high-yielding 

 tree will give rise to trees similar to the parent. This is hardly 

 likely to be the case under present plantation conditions, where 

 high and low-yielding trees are indiscriminately intermixed. But 

 it does seem to be indicated that if high-yielding trees can be 

 segregated, and provided that pollen of poor-yielding trees be 

 prevented access to the flowers, seeds capable of producing trees 

 yielding a high percentage of rubber would be assured. 



The author also gives particulars of the correlation between 

 yield and girth, which indicate that though there is a definite 

 correlation — trees with a large trunk being good yielders and 

 those with small trunks being poor yielders — it is not suffi- 

 ciently well indicated to be of great value in eliminating trees 

 from a plantation. 



It is of interest to notice that A." A. L. Rutgers {** Selectie en 

 Uitdunning/' Archief voor de Rubber culturer, 1919, 3, yv- 

 105-123) has made observations in Sumatra in full accord with 

 those of Mr. "Whitby and finds that " good trees remain good, 

 poor trees remain poor.' 5 a. w. h. 



