A phyto-geographical investigation in the Dutch East Indies has 

 to overcome a score of difficulties which in the temperate zone 

 either do nat present themselves at all, or, if they do, only to a very 

 slight extent. Complete modern handbooks enabling the non-specialist 

 to obtain the indispensable knowledge of the flora without spending 

 much valuable time, do not exist. This lack, the immense number 

 of species, and the great variability which many of these show, 

 create very serious difficulties to the phyto-geographer when naming 

 his plants and it is clear that a correct naming of the plants is one 

 of the fundamentals of phyto-geographical studies. The determination 

 of sterile material offers still greater difficulties. Unhappily, sterile 

 plants, as a rule, form the greater part of the materials the phyto- 

 geographer in the tropics has to work with. Everyone who has made 

 an extensive floristic investigation in the Dutch East Indian jungles, 

 has obtained the experience that, in a given locality, only a small 

 part of the vegetation bears flowers or fruits, a rule especially 

 applying to the shrubs and trees which generally predominate. Often 

 the trees are so high that materials from them cannot be obtained. 

 The comparative scarcity of flowers also causes that most not 

 flowering or not fruiting plants remain unobserved if one does not 

 examine the vegetation scrupulously; amidst the luxuriant green they 

 escape the eye. To a yet higher degree this rule holds good for 

 persons insufficiently acquainted with tropical forms; he who does 

 not know plants does not see them. Not small is, furthermore, the 

 number of plants which appear above-ground or.ly during part of the 

 year so that in localities where, at a given time, they may abound, 

 a few months afterwards not a single specimen is to be seen. Very 

 patient and minute investigations, reiterated in all seasons, are 

 required in order to draw any reliable conclusions. 



Another serious difficulty, not to be underestimated, is our lack 

 of knowledge regarding the relations between the tropical vegetation 

 and soil, climate and other oecological factors. At the present time, 

 now that soil and climate of Java are rather better known than formerly, 

 this difficulty might be partly removed by studying extensive and pro- 

 perly labelled collections. But both manner of labelling and complete- 

 ness of herbaria in the tropics leave, as a rule, much to be desired. 

 The following complaint of R i k I i ') is quite appropriate to many 

 of the collections in these herbaria: ,,Alle friiheren Forschungsrei- 

 senden bringen in ihren Sammlungen und Aufzeichnungen beinahe gar 



l ) Fortschritte Naturwiss. Forschung III (1911) 214. 



