ik 



M.. 



A FORTNIGHTLY REVIEW 



OF THE 



IMPERIAL DEPARTMENT OF AGRICULTURE FOR THE WEST INDIES. 





Vol. XV. No. 364. 



BARBADOS, APRIL S, 1916. 



Prick \d. 



CONTENTS. 



Page. 



Antigua, Re-afforestation in 128 

 British Cotton Growini; 

 Association 127 



117 

 118 

 12(1 



Page. 



Coco-nuts in Dominica ... 

 Cotton Conference: — 



Fir>t Day'.s Proceedings 

 Cotton Conference, We.st 



Indian 



Ceylon Botanic Gardens. 



Hi.story of 



Departmental Reports ... 



Disea.ses of Plants: — 

 The Interuid Hisease 

 of Cotton Bolls 



Durian Seeds, Experiment 

 in Shiijping 



Eucalyptu.s in Dominica ... 



Fibre and Cotton in tlie 

 Belgian Congo 



Field Experiments, Plan- 

 ning of 



Gleanings ... 



121 

 114 



12(i 



IKi 

 128 



... 121 



113 

 124 



Horses, Feeding Experi- 

 ments with Sugar and 

 Meat-Meal ... " 



Insect Notes: — 



Insect Pests in British 



Guiana 



The ISanana Weevil 



Market Reports 



Motor Plougli Demonstra- 

 tions in Englanil and 

 France 



Notes and Comments ... 



Papain from Montserrat... 



Philipliines, Imjirovenient 

 of Tropical Fruit in 

 the 



Rainfall in Dominica, IfHo 



Rape, Pa.sturing Pigs on... 



iSea Island Cotton Market 



Soil Exjilosives, Value of 



Vegetable Growing for 

 Canadian Markets 



Vitamines . 



117 



122 



12;; 



12S 



117 

 120 

 117 



ik; 



120 

 125 



11!) 

 127 



12.-> 

 121 



The Planning of Field Experiments. 



» 



.'HE bulk of the experimental work con- 

 ucted at Botanic Stations in the West 



^ Indies and elsewhere consists in plot trials 

 in connexion with plant selection, the testing of new 

 strains, the nianurial re()niieiiieiits of plants, and the 

 like. Until recent years this work was regarded as 

 being comparatively simple and straightforward, and 

 such as anyone familiar with i)lant culture might 

 arrange and carry out. It is now generally recognized 

 that such is not the case, at least as regards the plan- 

 ning and arrangement of such work. To be moderately 



certain that useful and reliable results shall accrue 

 from agricultural experiments, a considerable amount 

 of preliminary theoretical consideration is necessary; 

 one needs to be clear as to what it is one wishes to 

 demonstrate, and what degree of precision in methods 

 is necessary to solve the problem or problems proposed. 

 In other words, scientific methods are just as neces- 

 sary with field experiments as with laboratory experi- 

 ments. The chief difference between the two is that, 

 in laboratory experiments the conditions of the 

 experiments can be regulated, that is, the number of 

 factors can be strictly limited and controlled; in 

 field experiments there are a large number of niter- 

 acting factors which arc not under control, and our 

 results are largely subservient to the law of chances. 

 The only way in which we can hope to approxi- 

 mate to the truth, is by manifold repetition, by 

 the careful recordance of Huctuating factors of environ- 

 ment like rainfall and humidity, and by correlating 

 these with tht' results of experiments. 



Ideas as to the manner in which manurial experi- 

 ments should be conducted have of recent years under- 

 gone considerable modification. In the first place, 

 extensive duplication is now considered necessary. 

 tTenerally speaking, one control or untreated plot 

 and a series of untreated plots are not enough. They 

 ought to be duplicated several times according 

 to cii'cu instances. This is particularly important 

 as regards the control plot, for it is essential to he 

 able to compare the vaiiation between two controls, 

 as well as that between a control and a treated 

 plot. Manurial experiments with crops like sugar- 

 cane and provisions, in which the material produced 

 is simply vegetative growth, are of a more simple 

 character than manurial experiments with fruit-pro- 

 ducing crops like cacao and limes. With these crops. 



