78 AGRICULTURAL DEPARTMENTS AND RESEARCH 



on the experiment on the fields of some large experiment 

 plantation, or he must do it on a plantation in collabora- 

 tion with the manager of the plantation (or on a field of a 

 small labourer in collaboration with the owner). 



The last-mentioned method is to be preferred; it has 

 two great advantages : 



(1) In collaboration with a practical planter or native 

 proprietor, the methods can be investigated thoroughly 

 from a real practical point of view. 



(2) For the demonstration and propagation of a success- 

 ful new method an experiment on a plantation or a field 

 of a small proprietor is much more effective than one 

 carried out in an experiment garden or an experiment 

 plantation. 



A few examples may be given in support of this con- 

 tention : 



Experiments to fight the witch-broom disease of cocoa 

 in Surinam were started in 1904 and 1905 on three planta- 

 tions in collaboration with the managers. The method 

 applied consisted in removing the entire leaf-^bearing 

 crown, followed by spraying a rather drastic measure, 

 about which most of the planters were very sceptical. 

 After a few years experimenting, however, we knew not 

 only that the method successfully eradicated the disease, 

 but we knew also exactly what expenses for labour, tools, 

 etc., it required, and the practical difficulties the planter 

 had to be prepared to meet when he applied it himself. 

 Further, the neighbours and other planters came to the 

 plantation occasionally to follow the results, and when 

 we were convinced of the success of the method it was 

 unnecessary to begin a campaign for demonstrating, 

 propagating, or advertising the method. This result 

 could never have been obtained so quickly if the experi- 

 ment had been carried out in an experiment garden. 



It is quite true that not every planter and not every 

 small proprietor is fit for such a collaboration, and it is 

 an important part of the task of the scientist in the field 

 to select the right man. 



Breeding experiments with cocoa and coffee were com- 

 menced two years ago by one of the private experiment 

 stations in Java. The work was started on two cocoa 



