46 



owing to the fact that the damage they do is continuous and not 

 seasonal. In the Dutch East Indies the coconut palm is a very- 

 important economic factor in the life of the natives, but in spite of this 

 very little original work has been done there as regards its enemies. 



The work here described was begun at Buitenzorg (Java) and 

 continued at Padang, on the West Coast of Sumatra, where 0. rhinoceros 

 (usually associated with Rhynchophorus) does considerable damage. 

 Counts in various localities showed that 60 to 80 per cent, of all the 

 coconut palms were injured, and over 10 per cent, were dead. The 

 annual loss may be estimated at several million florins in the Dutch 

 East Indies. 



Particulars are given of the recorded world distribution of 

 O. rhinoceros. It is found throughout the Dutch East Indies, but does 

 not often occur at altitudes above 2,300 ft. Two other species, 

 0. tritubercnlatits, Lansb., and 0. centaiirus, Stemb., are more locally 

 distributed. A short description of 0. rhinoceros and 0. trituherculatus 

 is given. 



Previous investigations are briefly reviewed. The life-histor}'' of 

 O. rhinoceros was still imperfectly known, especially as regards the 

 duration of the larval stage, and consequently of the whole develop- 

 mental period, which latter has been variously estimated between 

 5 months (Friederichs) and 2 years (Banks). The whole subject 

 was therefore worked out independently. 



In the low-h'ing coastal districts the egg-stage averages 12 days. 

 In trap-heaps the eggs are usually found in the lower parts. Owing to 

 their white colour they are easily seen. 



The newly hatched larva is entirely white, but later on its head 

 becomes brown. The young larva is grey blue. In older or mature 

 larvae this colour turns to white or yellowish white. As there is a 

 risk of confusing the larvae of 0. rhinoceros with others that occur 

 in the same breeding places, a table is given showing the nature of the 

 breeding places and the differences distinguishing the larvae. The 

 length of the larval stage varied roughly from 2 to 4 months, 

 a result that differs materially from published ones ; the latter may 

 be due to poor feeding or other abnormal conditions. 



It has been stated that the larvae can live in marshy ground, but 

 there was no corroboration of this during these investigations. The 

 larvae occur only in soils composed of substances on which they are 

 able to feed, and they do not gnaw the roots of living plants. The 

 larvae found attacking roots are those of Xylotrnpes and other species. 

 The larvae of Oryctes do, however, use their powerful jaws for gnawing 

 wood that is not rotten. 



The following list of breeding places, in order of decreasing import- 

 ance, was compiled : — Communal refuse heaps of villages and towns, 

 nipa-palm rubbish, dying and dead (erect or fallen) trunks of coconut 

 and other palms {Oncosperma horrida, Corypha gebanga, Metroxylon 

 sagus, Elaeis guineensis) , vegetable rubbish (coffee and cacao husks, 

 leaves and fruit peel, etc.), decayed dung of cattle, buffalos and horses, 

 rubbish from sugar-cane and rice-husking mills. The author has not 

 found 0. rhinoceros in the rotten stems of trees other than palms, as 

 stated by Friederichs, but Corporaal has done so, though this appears 

 to be exceptional. Examination showed that vegetable rubbish that 

 is one month old is suitable for oviposition and larval development. 

 In small quantities of refuse, e.g. about a cubic yard, the attraction 

 for the beetles reaches its maximum after 3 months and then 

 rapidly decreases. In 8-9 months the refuse has turned to 



