EEVISION OF THE JAPANESE TEEMITES. 



17 



plantula. Hind legs exceeding the end of abdomen. — (Text-fig. 

 1, k). 



Abdomen. — Long, with a high dorsal arch. Cerci long, hairy, 

 obscurely 3 -jointed ; the last segment slender, but a little longer 

 than the two preceding ones combined. Styli very long, longer 

 than the last joint of the cerci, with a few hairs at the tip. — 

 (Text-fig. 1, 1). 



Measurements": — 





Length (in m.m.) 



Breadth (in m.m.) 



Body 



16—19 





Head 



4.35—6.05 



3.9—4.7 



Head with mandibles 



6.7 —8.2 





Antenna (25-joints) 



5.9 





Mandible (right) 



3.0 —3.5 





Labrum 



0.8 —0.9 



1.1—1.3 



1) In order to avoid confusion, I may say a few words here on the method of taking 

 measurements of the -various parts of termite, adopted in the present paper. 



In most cases as many individuals as possible were measured. Larger parts such as the 

 body, wing, etc. were measured by ordinary method, whereas each of minute parts such as the 

 head, antenna, labrum, pronotum, etc. was isolated from others and measured under a microscope. 



By the length of body is meant that between the foremost portion of the head (this must 

 be either the anterior end of the mandible or that of the labrum, as the case may be) to the end 

 of the abdomen. 



By the length of head is meant that distance measured from the anterior margin of ante- 

 clypeus to the posterior margin of the head. In the case of soldiers of Eutermes, this is from 

 the tip of the snout to posterior margin of the head. 



The breadth of head is taken at its broadest portion, including the eyes. In the case of 

 soldiers of Eutermes the length of snout is from its tip to the line connecting the anterior 

 margins of both antennal clefts. 



The length of labrum is from its anterior margin to the line, which connects its postero- 

 lateral corners. 



The length of wing is from the cross-sture to its distal end. 



As to other parts such as pronotum, mesonotum, etc. it needs hardly be mentioned that 

 lengths are taken along their median lines, and breadth, across the broadest portions. 



