148 



THE AGRICULTURAL NTIWS. 



May 6, 1916. 



ORCHARD CROPS. 



COCO-NUT SELECTION. 



Some u-etul selection work on coco-niits is being 

 conducted in Fiji, it being considered that improvement 

 in the yielding capacity of these trees is more likely to 

 be obtained in this way th>in in any other. The 

 following consists of extracts taken from a Bulletin on 

 the subject (No. >i), issued by the J)epartment of 

 Agriculture, Fiji. 



SELECTION. 



Sixteen palms were selected in the coco-nut blocks on 

 Vuua e.state, Taviuni, and one nut from each tree will be 

 planted in each plot so that all the plots will be alike, as far 

 as the origin of the seed goes. This is regarded as an 

 important point, as it will be possible to trace whether 

 the results of the seed from any one pahn are due to local 

 conditions or to the characters of the plant. The whole series 

 of plots will also show to what extent coco-nuts keep true 

 to type. 



In selecting the parent trees, the aim has been to get the 

 palms of .several distinct and well-marked varieties which 

 have been observed to be good bearers. Other well defined 

 varieties are to be met with, but as they were not known to 

 be good bearers, or were in some cases known to be poor 

 bearers, they have been avoided. 



.\ knowledge of the average results of the pilms, from 

 which the selection was made, over a series of years was 

 e.ssential, and on this point the advice given by Mr. H. V. Tarte, 

 of the Vuna estate, Taviuni, was invaluable, so that there can 

 be no doubt that the selected palms are all good bearers. 



The points usually recommended for the selection of 

 coco-nuts have been amended in accordance with suggestions of 

 Mr. H. V. Tarte, whose observations as to the bearing powers 

 of varieties of coco-nuts are of the greatest value. 



The chief points to be considered are of course the 

 amount of kernel in each nut, and the number of nuts borne 

 each year. Roth these characters are found to be associated 

 with certain features more readily observed by insjiection, and 

 it is upon these 'outside' features that the selection \i based. 



Besides any special characters, the tree must be in 

 a healthy condition and of vigorous appearance with deep 

 green leaflets. The leaf-stalks may be either green or 

 have a red colour of varying degrees of intensity, the 

 mature nuts having the same colour as the leaf -stalks. 

 Palms with light green leaflets should always be 

 avoided. The age of the tree is not considered very 

 important so long as the palm is bearing well. The 

 plants shi'uld be taken from the general cultivation 

 rather than from some particular spot where the con- 

 ditions may be considerably more favourable than the 

 average. The plants should have a good 'head'. Sometimes 

 leaves are to be seen occupying only the upper half of the 

 sphere of which the 'cabbage' is the centre, and the ti[)s of the 

 leaves the circumference. In other cases leaves occupy 

 practically the whole of tlu> sf)here. In the former case the 

 efTect would be caused by the leaves dropping off at a com- 

 paratively early stage, while in the latter case they persist 

 for a long time. Since the leaves are responsible for the 

 ■work of converting the material taken up by the roots and 

 from the air into material suitable for use by the giowing 

 parts of the plant, it is reasonable to suiipo.se that where 

 a tree carries a better 'head' of leaves it will give a better 



crop. It may be argued that such trees are more likely to 

 damage by storm; this is, however, a risk that must be 

 taken. Those palms having slender stems are to be preferred, 

 as they are found to give more to the wind and so escape 

 being broken oH or blown over as may be the case with the 

 unyielding stout tree. The tree must be satisfactory as 

 regards health and vigour, otherwise the slenderness may be 

 a sign of poor nutrition. The stem should be uniform and 

 not rapidly tapering. 



Some variation is seen in the length of the stalk of the 

 bunch. It is not, certain whether this is a permanent 

 character or not, but observation has shown that plants 

 showing this character are often good bearers. 



The nuts should be nan ow at the stalk end. Nuts wide 

 at this point tend to squeeze one another out of the bunch. 

 In order to compare one strain with another as regards this 

 point, the method used was to measure the distance between 

 two blocks, 1 inch high, the edges of which just touched 

 the nut placed upright with the stalk end on a table. The 

 husk should be thin. Whatever variety is being chosen, 

 medium-sized nuts should be .selected. 



A table is here given in the Bulletin showing the 

 characters of the varieties selected as regards (a) age of 

 tree, (b) girth at -5 feet, (c) colour of mature nuts, (d) 

 girth of nut, (e) thickness of husk, (f) length, (g) weight 

 of dry nuts, (-h ) weight of water, (i) weight of kernel, (k) 

 calculated weight of copra, (1) calculated ntnnber of nuts 

 per ton of copra. The measurements for individual 

 nuts of one and the same variety or strain exhibited 

 remarkable uniformity. The calculated number of 

 nuts to make a ton of copra showed a variation of 

 4,97<S and 3,381 in the sixteen averaged results put 

 forward. 



CoI'RA' FROM SA.MPLK SEED-XUTS. 



The kernels of the first eight varieties were exposed to 

 the sun with the idea of preparing copra, but two rainy days 

 set in and the copra became ver\' mouldy. In addition to 

 this there was some uncertainty as to whether some of the 

 copra from two jiarticular nuts had not been mixed together. 

 The total weight of copra amounted to 57"1 per cent, of the 



kernels The weather for five days was fine and 



dry with fairly strong .south-east winds; the kernels were not 

 exposed tf) the sun at all. Drying proceeded very well, and 

 during that t;ime .'57 per cent, of the kernels dried out. In 

 five more days excellent samples of copra were obtained from 

 all the kernels, no signs of mould being seen. Each lot of 

 copra was carefully weighed, the percentages of copra in the 

 kernel being from -^S'S to 51 '3, the average beins 

 59!). 



It may be remarked that with 3,S54 nuts to mike 1 ton 

 of copra or 19;? to make 1 cwt., four nuts per tree per annum 

 will give 1 cwt. of copra per acre with trees set 30 feet by 

 .'50 feet apart. 



In considering this point it must be borne in mind that 

 the figures are obtained from si.vteen nuts only; they were, 

 however, the nuts which came nearest to the average measure- 

 ment of each kind; the kernels were most Ciirefully weighed 

 to the nearest gram (•0022 of IB).). It is true that they were 

 selected nuts to start with, but as the improvement of coco- 

 nut cultivation, as in most other agricultural industries, 

 depends largely on careful selection, the figures may be taken 

 as indicating what should be jwssible. 



