496 



required only about 193 days, with a larval period of 159. This specimen, 

 in which the resting larval period was either very short or altogether 

 absent, completed development at the end of the rainy season, while the 

 eleven others did so before the beginning of the next rainy season, 

 passing their resting larval period in the colder dry season. The time 

 for the single specimen referred to above agrees to some extent with 

 Leefmans' results, while that for the eleven others agrees with the 

 period established for British India by Ghosh. 



The results in Sumatra and Samoa are compared. Leefmans worked 

 out a means of poisoning trap heaps with arsenic, and his method of 

 preventing oviposition by means of a layer of sand eliminates most of 

 the breeding-places (almost always rotting palm wood and rarely 

 palm stumps) in the Dutch East Indies. In Samoa the position is 

 quite different. Few or no species of trees are rejected for oviposition. 

 Any tree stumps are likely to be infested, so that new fellings are a great 

 danger to palms and the employment of sand is useless. In Sumatra the 

 breeding-places are chosen near palms, so that the spread of the pest is 

 limited. The same thing happens in Samoa, but palms there grow 

 nearly everywhere, and the bush palm, Cyphokentia samoensis, is found 

 in the forests near coconut plantations. In some parts, too, wind- 

 carriage is an important factor. Consequently in Samoa 0. rhinoceros 

 has a much greater variety of breeding-places and is far more difficult 

 to reach than in the Dutch East Indies. The authors accept Leefmans' 

 view that driftwood affords a means of spread. 



In Sumatra no use has been made of M etarrhizium anisopliae. The 

 successful results in Samoa with this fungus were due to the intensifica- 

 tion of an existing infection and not to the creation of one. About three 

 months after a trap-heap has been established it is opened and all the 

 adult beetles are destroyed. The larvae and eggs are mixed with fungus 

 material, which is used at the rate of 10 dcm. per cubic metre of trap 

 heap material [about 610 cu. in. per 35 cu. ft.], and are buried at the 

 bottom of the heap, which is then remade. In a small heap 100 larvae 

 are considered enough, and an excess of larvae from one heap is used to 

 make up any deficiency in another. About six weeks afterwards the 

 heap is examined and then closed up without removing anything. 

 Three months later a third examination is made, and the fungus material 

 is renewed if necessary. The author knows of no case where the infection 

 has failed. 



An investigation of the extent to which 0. rhinoceros penetrates into 

 virgin forest adjoining a coconut plantation showed that only the edges 

 are involved, no trace of the beetle being found farther than about 

 40 yards. The case may be different in forests where many bush 

 palms are growing. In 1916 the junior author noted that bush palms 

 are attacked only after neighbouring coconut palms have been 

 destroyed. 



Data are given on the extent of infestation before the war and up 

 to 1916. In the latter year the position was about the same as before. 

 In newly-infested localities there was a large increase of injury, but 

 in old-infested places no increase was noted. The devastating stage 

 of the infestation was already over. Conditions in the first half of 

 1920 were roughly the same as in 1916. 



Present measures are of an administrative character, such as 

 collection and breeding- place elimination. The latter is perhaps done 

 more thoroughly than in pre-war times, but this is outweighed by the 

 entire discontinuance of biological measures. To reduce the infestation 



