on the Habits of ScolytidcB and Bostriclddce . 219 



a large tree in the Aru Islands, of a kind containing abundance 

 of milky sap, which hardened on exposure to air very much like 

 " gutta percha." A few days after I found on it dozens of 

 a species of Scolytidce, with their abdomens protruding from the 

 holes they had bored, but all dead. With a remarkable deficiency 

 both of instinct and reason, the little creatures had dug their own 

 graves, and were all glued fast by the hardening of the milky 

 sap. In a few days more there were hundreds so killed ; indeed 

 it appeared as if not one escaped. It seems evident, therefore, 

 that this tree coidd not have been the proper food of this species, 

 or the right place to deposit its eggs. I have since observed 

 exactly the same occurrence in another locality. 



In my hut in the Macassar forest, built for me of freshly cut 

 bamboos, palms and timber, the Scolyti, S^c. literally swarmed. 

 In a little more than a month, several pints of wood dust, as fine 

 as flour, had accumulated on the foot of the posts and on some of 

 the beams. In the stillness of evening their never ceasing jaws 

 could plainly be heard at work, producing a slight sharp creak. 

 In this house I took almost all my Macassar species. 



Now these facts lead us to conclude that the Bostrichi and 

 Scolyti attack only dead wood, generally in the first stage of 

 drying or decay ; for if their proper and usual food was living 

 wood, why should they all rush as to a feast directly a tree is cut 

 and begins to dry ? 



It may be said there are other species that attack living trees, 

 but the negative evidence is very strong against such a suppo- 

 sition, which is besides altogether a gratuitous and unnecessary 

 assumption if not supported by direct evidence. In five years' 

 search after insects in the eastern forests I have never met with 

 one, and the thousand sharp eyes of European Entomologists do 

 not appear to have been more successful. This is the more ex- 

 traordinary, as it is evident that a tree cannot be injured or killed 

 in a moment ; weeks, perhaps months, would be required before 

 any part of the wood or bark would become actually dead, and 

 during all this time the little round holes that the insects bore, 

 and the wood-dust that has no other exit, must be easily dis- 

 covered. 



I am led, therefore, to conclude that the Scolyti, &c. attack 

 wood in which the vital forces have ceased to act ; and they are 

 able to detect this before any external change has taken place. 

 It is only at a later period that we observe the tree to be suffering, 

 and in the parts most affected we discover the Scolyti to have been 

 at work, and erroneously impute the mischief to them. As well 



q2 



