KEYISIOX OF iSTEARCTIC TERMITES. 5 



Buckley was the first to write about the habits of our species. 

 Later Hagen. Hubbard, Schwarz, Joutel, and Heath studied the 

 biology of various species, and in recent years the junior author has 

 made detailed studies on several of oiu- species. 



CLASSIFICATION. 



The early di'. i-ions of Termes were ])ased on adult structui'es and 

 outline the groups now known as subfamilies. Wasmann in 1897 

 and later othere have j)laced much importance on the soldier as a 

 basis of division. The soldier is a liighly adaptive form and normally 

 in no way connected with reproduction and intercalated in the 

 normal plan of develo]^ment; it is therefore of little value in determin- 

 ing affinity. To the mature winged insect is intrusted the future 

 of the species. It exists in the winged state but a brief period; it 

 is therefore but little affected by environment, and retains more of 

 those minute ancestral characters which indicate relationship. In 

 view of the divei-se. conflicting systems proposed by different writers, 

 too much stress can not be placed upon the adult winged insect as 

 the sole source of classification. The winged form is the ] arent of 

 the winged form and the parent of the soldier, so that in it is found 

 the origin of all character of adult as well as those of the soldier; so 

 if the soldier possesses any structures of family or generic value 

 they are transmitted b}^ the winged form, which therefore is the 

 real source of these structures. 



The best cliaracters for group classification are found in the wings, 

 head, legs, and cerci. Most of the characters I have used have been 

 used in some sense before, but I have added as a family character 

 to distinguish the Termitinae the cilia on the wing margin. Tiiis is 

 correlated with a peculiar color of the wing. By this certain genera 

 are placed differentl}^ than by Holmgren, Init I feel certain that 

 these hairs are of far greater value than the shape of the pronotum, 

 which is far from being as constant as his tables would indicate. I 

 consider the group to be of two families, and our fauna divisible 

 into four subfamilies. The names of the families, according to all 

 codes, must be based on a contained genus; therefore I can not 

 accept the names given by Holmgren. 



The best generic classification, but unfortunatel.y v.^ithout generic 

 nam.es, was proposed by Haviland; he found good characters for 

 various groups based on both adult and soldier. In recent years 

 Nils Holmgren has published several large v/orks on the termites 

 of the world, more or less of a preliminary sort. Between his families 

 Meso and Metatermitidae there is no real distinguishing structure, 

 and the subfamilies are hardly better defined, the host of genera being 

 based sometimes on adults, sometimes on soldiers. However, he 

 has the immense advantage of studying large collections from all 

 parts of the world. 



