114 



THE AGltlCULTURAL NEWS 



April 8, 1916. 



Tvhich are permanent, we are frequentl}' faced with the 



manifestation of distinct individuality. If plots are 



selected at random and judged merely by general 



appearances, serious error may be introduced. The 



control plot, for instance, may contain a larger 



number of trees which are naturally heavier 



yielders than the treated plot — a circumstance that 



would tend to vitiate the effects of manurial treatment. 



Where manurial trials are conducted to determine the 



•effect of fertilizers on fruit yield, the natural yields of 



the trees should be recorded first. The effect of 



manures on developing trees before they come into 



bearing, is another matter and concerns vegetative 



growth. It has to be remembered, foo, that in manuring 



■orchard crops the effect of manures on the soil as 



well as on the plant is introduced. With annual crops 



where thorough cultivation is practiced periodically 



together with rotation of crops, this effect is not felt to 



the full extent. 



- ■*+/ ■ 

 From year to year a large number of plot trials 



Tiave been made in the West Indies with different 

 varieties of provision crops. It would appear that in 

 some instances, as with some of the experiments with 

 varieties of sweet potato, the results obtained are not 

 •conclusive. It has been the custom to conduct each 

 year, in single plots of very small area, a trial of each 

 rariety. The mean of several years' results with each 

 variety is regarded as representing the relative pro- 

 ductive power of that variety. It would seem that a 

 •quicker and more reliable insight into their relative 

 productive powers could be obtained ifinstead of .spread- 

 ing the experiment over a number of years, all the trials 

 were conducted in one year, each plot being duplicated 

 many times. The mean of ten years' results has not 

 the same scientific significance as the mean of ten 

 results obtained in one year. In referring to these 

 experiments, however, it is only fiiir to point out that, 

 apart fiom their object of showing the relative values 

 of different varieties, they are the source of an annual 

 .supply of planting material for distribution amongst 

 peasant grovt'ers; and so the work has another object, 

 the recording of yields being partly incidental. 



The selection of seed is another line of experi- 

 mental activity which calls for preliminary thought 

 and consideration. This has been recogruzed in 

 -connexion with cotton, as is shown by the success 

 obtained by means of individual plant rather than 

 mass selection. Attention is only just being given to 

 the selection of the coco-nut. Here again we must go 

 to the individual plant, though not in quite the same 

 sense as in the case of cotton selection. Nuts should 

 ))e selected from the most productive trees growing 



under the same conditions as the less productive ones. 

 The kind of nut to be selected will depend upon what 

 section of the market for coco-nut products one intends 

 to supply. If copra is to be the principal product, 

 then large nuts with a high endosperm content will be 

 selected. Where large numbers of the nuts themselves 

 are shipped, the aim of the' grower will be to produce 

 uniform nuts juSt above the standard size. 



All field experiments are influenced to a very 

 large extent by , climatic conditions principally relating 

 to moisture. It is therefore important that all field 

 experiments should be accompanied by systematic 

 weather observations. Hitherto the rain-gauge has 

 been considered the only essential equipment. As 

 a matter of fact the bare recording of the precipita- 

 tion of rain is almost valueless. It is necessary to 

 know, as well, at what times the rain fell, so that the 

 effective rainfall can be calculated. Of even greater 

 importance are humidity observations, since humidity 

 is closely correlated with growth. Wind, being 

 inimical to growth, shou]d also be recorded by an 

 anemometer. In cases where it is intended to caiTy 

 on plot experiments from one year to another, these 

 climatic observations are especially important as 

 affording about the only means of arriving at the cause 

 or part cause of variation in annual results. 



LEEWAHD ISLANDS: REPORT ON EXPERI- 

 MENTS WITH SUGAR-CAN K CONDUCTED IN 

 ANTKIUA AND ST. KIYTS, 191J,-lo. 



This report is divided, as usual, into two parts, one 

 concerning varictie.s of sugar-cane, the other manurial 

 experiments with .sugar-cane. It may be thought, and with 

 some ju.stitication, that the publication of this report which 

 deals with the year ending March 31, 1915, is somewhat 

 belated; it should be borne in mind, however, that a summary 

 of the results is placed before the Local Agricultural 

 Societies in the two islands immediately the results have been 

 got out. Tims this report is intended to serve purposes of 

 record, and to enable workers in other parts of the world 

 to follow what is happening in the Leeward Islands. 



r.vuT I — Experiments w rni V.mueties. 



The experiments with varieties of canes are dealt with 

 first under the heading Tlant Canes in Antigua'. The mean 

 return from all stations for the year under review shows 



