TROPICAL FLORAS 85 



to the denudation of the rich surface soil bj torrential rains ; 

 this soil has been produced by countless ages of forest growth, 

 and it will require an equal lapse of time to reproduce it. 



Returning now to the more direct teachings of small areas 

 when methodically studied, I may add that Dr. Koorders has 

 informed me that some years since he made a visit to 

 Minahassa, in X. Celebes, and in four months, between the 

 sea-level and 6500 feet, he collected or observed about 2000 

 species of flowering plants, of wdiich about 700 were forest 

 trees. As these last are Dr. Koorders' special study it is 

 to be presumed he paid great attention to them, yet he could 

 hardly have obtained such a complete knowledge of them in 

 a few months as in the ^' reserves " of Java, where, in suc- 

 cessive years, not a single species could have escaped dis- 

 covery. This would imply that the forest flora of Xorth 

 Celebes is even richer than that of Java, and it is almost 

 certainly more peculiar. And if the larger islands of the 

 Moluccas — Gilolo, Batchian, and Ceram — are equally rich 

 (and they have all the appearance of being so), then every 

 estimate yet made of the species-population of the whole 

 Archipelago must be very far below the actual numbers. 



There must be hundreds of young botanists in Europe and 

 America who w^ould be glad to go to collect, say for three 

 years, in any of these islands if their expenses were paid. 

 There would be work for fifty of them, and if they were prop- 

 erly distributed over the islands from Sumatra to Xew Guinea 

 in places decided upon by a committee of botanists who knew 

 the country, with instructions to limit their work to a small 

 area which they could examine thoroughly, to make forest 

 trees their main object, but obtain all other flowering plants 

 they met with, a more thorough and useful botanical explora- 

 tion Avould be the result than the labours of all other col- 

 lectors in the same area have accomplished, or are likely to 

 accomplish, during the next century. And if each of these 

 collectors had a moderate salary for another three years in 

 order to describe and publi^^h the results of their combined 



