OBSERVATIONS ON TERMITES IN JAMAICA 217 



touch the alien with antennae and then instantly, or after a 

 longer interval, to seize it with the mandibles either violently 

 or mildly. The mandibles often engage the mouth parts of a 

 foe and the two stand with locked jaws till after some time 

 one may have its mandible bitten off or be injured by loss of 

 blood. Again a frequent point of attack is along side where 

 one of the legs is seized by the knee and bitten till it separates, 

 then a second leg may be cut off at the knee. A third place of 

 attack is the anterior or the posterior part of the abdomen 

 and a long continued biting there results in the loss of blood 

 and death with shriveling of the body. An introduced alien 

 commonly becomes the centre of interest of several or many 

 termites which surround and examine it and 5 or 6 may at 

 once seize it. In this way an introduced worker was lifted out 

 with several alien workers clinging to it, biting its legs or cut- 

 ting into its abdomen, and not removed without tearing the in- 

 troduced worker to pieces. 



When an alien runs about it may be followed and overtaken 

 by attacking termites, and if lifted out the attacking termites 

 seem to search for it for a short time. 



The method of attack and defense shown by the soldier was 

 to point the head at the foe, touching it or nearly touching it 

 with the tip of its snout-like gland nozzle and then to eject 

 minute drops of clear liquid onto the head of the foe. The drops 

 were sometimes seen at the tip of the snout and then on the 

 head of the foe, but more often the secretion first became visible 

 as very minute droplets on the head of the foe. Over the head 

 of the foe these droplets sometimes lay in lines as if squirted 

 some distance through the air. This secretion seemed to daze 

 the foe or to inconvenience its movements ; sometimes this was 

 visibly connected with the sticking of the antennae to the head. 

 In this way the antennae might become so adherent that they 

 would break rather than come loose when the loop was pulled 

 with a needle. The drops adhered to vegetable fibre and could 

 be pulled out as a viscid substance. 



Returning to the consideration of the interactions of entire 

 communities as indicated in the bowl experiments, we first 

 noticed exception to the general hostility of termites to aliens 

 in the case of two communities taken from trees a quarter of 

 a mile apart, which we will call communities i and 2. Members 



