Bramble-Dwellers 



perlment, the strings of cocoons are arranged 

 differently in different tubes. In some of them, 

 the heads of the cocoons are turned down- 

 wards, towards the opening; in others, they 

 are turned upwards, towards the closed end; 

 in others again, the cocoons alternate in direc- 

 tion, that is to say, they are placed head to 

 head and rear to rear, turn and turn about. 

 I need not say that the separating floors are 

 of sorghum. 



The result is identical in all these tubes. If 

 the OsmiiE have their heads pointing upwards, 

 they attack the partition above them, as hap- 

 pens under normal conditions; if their heads 

 point downwards, they turn round in their 

 cells and set to work as usual. In short, the 

 general outward trend is towards the top, in 

 whatever position the cocoon be placed. 



We here see manifestly at work the in- 

 fluence of gravity, which warns the insect of 

 its reversed position and makes it turn round, 

 even as it would warn us if we ourselves hap- 

 pened to be hanging head downwards. In 

 natural conditions, the insect has but to follow 

 the counsels of gravity, which tells It to dig 

 upwards, and it will Infallibly reach the exit- 

 door situated at the upper end. But, in my 



36 



