668 termites or white ants. 



for defense. The segments of the antennae are more elongate than in 

 the males and females, and fewer, generally in the proportion of 

 8 to 9. The mandibles arc very various in the different species, but very 

 characteristic of each species, and quite different from those in the 

 males and workers. The gula is large and firmly united to the head, 

 generally for the greater portion of its length. The cervical sclerites 

 are larger than in the males and workers. The thorax and abdomen 

 are generally but little chitinized. The latter is generally more 

 quadrate than in the workers. Some individuals have rudiments of 

 ovaries, and some of testes; but the ventral plates of the 8th and 9th 

 abdominal segments are always entire. 



The workers are wingless and for the most part blind; they are but 

 little chitinized. and larval in appearance. The head is round, the an- 

 tenna' are shorter than in either male or soldier, and the number of 

 segments intermediate. The mandibles are short and powerful and 

 covered by the obtuse labrum. In species which nest in the wood on 

 which they live the form is cylindrical, and the legs shorter than the 

 abdomen In species which wander much in search of food the thorax 

 is considerably narrowed, and the legs longer than the abdomen. 



Termites inhabit all the warm regions of the earth in countless num- 

 bers. They are unable to withstand a prolonged winter's frost. Their 

 greatest enemies are ants. Their chief means of defense is their power 

 of burrowing and building. 



CLASSIFICATION. 



In the matter of genera I have followed Hagen. His genera admit 

 of distinctions common to every caste. The genus Termes contains 

 numerous species of very diverse forms and habits, yet it can not be 

 subdivided by characters common to every caste. * * * 



The genus Termes is so large that Hagen, who tried to make several 

 genera of it, failed owing to the incompleteness of his material. I 

 also have failed, and think that in the interests of naturalists the 

 attempt should be postponed. The genus does, however, present nat- 

 ural groups and these I have attempted to define, but more material 

 and further examination will alter the definitions and limits I have 

 given. The groups can seldom be distinguished by characters common 

 to every caste, nor are the limits of the groups the same if we rely on 

 the soldiers as if we rely on the males. 



The largest forms of the genus are fungus growers. There is an 

 American group of large termites, represented by T. dirus, which are 

 almost certainly fungus growers; the soldiers have a pair of lateral 

 horizontal spines on the pronotum. There are three Old World groups 

 of fungus growers. The most important is represented by T. IxUi- 

 cosus; it builds tall mounds; the imago and soldiers are of large size. 



