78 Treubia Vol. HI, l. 



Further, two species of fresh-water Molluscs have been found and 

 a brackish-water species, which was also noticed in the lake on 

 Verlaten I. 



The number of Vermes observed on both islands is again the same, 

 consisting on each island of species belonging to the same groups. Earth- 

 worms are more abundant on Sebesy than on Krakatau. 



To obtain an answer as to the question whether the fauna of the 

 Krakatau islands was still abnormal, I also used another method. 



Jacobson has already stated in his paper that on Krakatau a limited 

 number of species occur, but that of most species there are plenty of indi- 

 viduals. The species which had newly invaded the island could increase 

 abnormally, their parasites and enemies having not yet reached the island. 



I searched for a method to put this thesis into figures and tried it by 

 examining how many animals, species and individuals, occur on one square 

 Meter. Only animals living under dead leaves and in the vegetable mould 

 immediately beneath them were counted. This surface layer of the soil was 

 chosen because in a tropical wood it is the only environment with uniform 

 conditions throughout a large area. 



The method was practised as follows: — 



On a certain spot where the soil was covered by a uniform layer of 

 vegetable debris, one square Meter was laid out. At this place the leaves 

 and mould were heaped up, the mould only so far as it could be easily 

 gathered together without loosening the soil. 



The decaying vegetable material was further sifted through a 5 mm. — 

 mesh wire sieve into a bag. The coarser material not going through the 

 sieve was searched on the spot to collect the larger animals. What went 

 through was taken along for a stricter examination, which took place the 

 same day whenever possible. The sifted material was spread out in a thin 

 layer on a white cloth and the animals were picked up by forceps, the 

 smallest ones by a fine hair brush. 



The method as practised by me does not give exact figures, only a 

 part of all the animals occurring on the examined place being caught and 

 many of the smaller ones being overlooked. 



Besides, soft species are often damaged by the sifting; these mutilated 

 animals covered with dust and sand are hard to detect, especially so if they 

 do not move or are already dead, in this manner chiefly soft larvae, Aptera 

 and woodlice are damaged ; the number of these animals given in the lists 

 is therefore proportionately always too low. 



But notwithstanding these drawbacks, the method used is, I believe, 

 the best to give in a short time comparatively reliable figures. For our 

 purpose we do not require the exact numbers but only relative ones. 



Occasionally time was not available to count all the individuals of 

 certain species, in which case only the species are listed, not the number 

 of specimens. Of ants always the species only are recorded. 



