362 MB- &• I>- HATILAND ON TERMITES. 



are broken off, for tlie apical segment is of a different shape from 

 the others. Although the segments o£ the antennae are fewer in 

 the soldier than in the male, they are generally longer and more 

 cylindrical, so that the antennae of the soldiers are often as long as 

 or longer than those of the imago. The antennae of the workers, 

 on the other hand, are always much shorter, yet the number of 

 segments which compose them is never less than in the soldier 

 and never more than in the male. The actual length of the 

 antennae in the genus Termes seems to be but little correlated 

 with the actual number of segments which compose them, 

 whether we compare the different species, or whether we compare 

 the different castes. Long antennae go with long legs, and this 

 is true w^hether we compare caste or species. Long legs and 

 long antennae go with much walking and foraging ; and this is 

 true when we look to differences between species, but not when 

 we look to differences between castes. Soldiers with long slender 

 legs belong to species which forage for food at a distance from 

 the nest : soldiers with short stout legs belong to species 

 sluggish in their movements, and which venture but little from 

 home. 



Blindness amongst the soldiers and workers is more universal 

 than it is in ants. There seems no reason to doubt that the 

 blindness is connected with the mode of life. The impossibility 

 of attributing the blindness to the inherited effects of disuse, 

 seeing that none of the parents in any of the species are blind, 

 utterly discredits such an explanation in the case of other blind 

 animals. 



In all the castes the abdomen varies greatly in size and 

 appearance according to the nature of its contents. 



The winged images have an unconquerable desire to leave the 

 nest, and to run the risk of dangers from which not one in many 

 thousands escapes. By this means it is that interbreeding and 

 distribution are effected. Dr. Fritz Miiller aptly compared the 

 winged individuals to perfect flowers, and the neoteiuic indi- 

 viduals to cleistogamic flowers. The comparison may be carried 

 a step further. In temperate chmates the winged forms appear 

 in early summer ; in equatorial regions they appear for the most 

 part in simultaneous swarms at favourable seasons, whilst in 

 some species they seem to be constantly produced in small 

 numbers the whole year round. The problems of when to 

 swarm and how many images to produce seem to be solved in 



