16 



The workers are usually short and broad in proportion to the soldiers, 

 have the head rounded, with similar antenna? to the winged form, but 

 without eyes or ocelli, the large powerful jaws, which are responsible for 

 all the damage caused b}^ the white ant, hidden when viewed from above 

 by the rounded upper lip. The thorax is short, with stout legs, and 

 no traces of wings, with the segmented abdomen somewhat curved on the 

 sides to the rounded tip. The soldier has the general appearance of the 

 worker, but the head is usually larger or more elongated, with the jaws 

 produced in front of the labrum or upper lip into a pair of sabre, sickle, 

 or tooth shear-bladed jaws, which meet at the tips or can be folded over 

 each other. While the legs and thorax are similar to that of the worker, 

 the abdomen is usually smaller and more slender in proportion. The 

 soldiers are sometimes almost as numerous in the nest and colonies as the 

 workers ; but in others you may tind hundreds of workers to every soldier, 

 probably because less protection is requii'ed in the underground nests. 

 The soldiers act as the police or protectors of the colony, rushing from alt 

 quarters as soon as a break occurs, and holding the breach, until the 

 workers can repair it with fresh clay, which they bring from the interior. 

 It is on the workers that tlic whole economy depends : they tend the queen, 

 remove and look after the eggs, feed the helpless pupie, and do all the 

 building and construction of the termitariuni. 



Though the queen is apparently the head-centre and sole source of life 

 and reproduction of the countless multitude that inhabit the nest, we 

 sometimes come across greatly-developed pupje produced from the em- 

 bryonic winged forms, though the wing pads have even become aborted, 

 which for want of a better name, we call supplementary queens. I had 

 twentj'-two specimens sent me taken from a small nest in southern 

 Queensland, and have frequently found two or three in a nest. 



These supplemeutar}' queens have Ix-en noticed in other countries, and 

 Dr. Miiller, when working out the life histories of the white ants at Santa 

 Catherina, Brazil, obtained thirty-one specimens in a nest. 



FolloAving up the relationship of the Tcrinitulrc with the order Orthop- 

 tera, I think if we take the Blattidce (cockroach) as the end of the order, 

 followed by the termites and Emhiidcc, they come very naturally together 

 as two more families in the order, without running them into the Pseudo- 

 Neuroptera. 



Hagen, in his monograijh, based his classification of the family upon 

 the structure of the wings, ocelli, number of joints in the antennae, shape 

 of thorax, and tibial spurs, which I followed in my work on the Aus- 

 tralian species. While the structure of the wing forms is very distinctive 

 in defining generic species, it will be found that the study of the soldiers 

 is one of the best methods of defining the species, and these differences 

 are constant ; when sometimes it is very hard to point out the difference 

 between two species in the perfect-winged form, even when the soldiers 

 can be readily distinguished. In dealing with the Australian species I 



