298 Annals of the South African Museum. 



o o Giebelia inirabilix ? . 



6 6 Giebelia he.rakon ? . 



Mackayia diTtwrpha s . 



Mackayia heteracanthus 



6 6 Mackayia heteracanthus ? . 



* 



In the above o == peg-like spine, 6 == do. on special branch of 

 antennal band, and . short spine. 



The marginal bands have been used in discriminating between the 

 ? ? . If the specimen of this sex of G. mirabilis supplied by 

 Kellogg is thoroughly typical it may be set aside at once by the 

 simplicity of the bands on segments 2 and 3. The $ 2 of the 

 4 spp. do not quite agree in these bands, and some of these are 

 figured for comparison. 



The 3 eye of the 4 species shows minute modifications which 

 we think of some importance. In liexakon (which is probably the 

 most primitive of all) and mirabilis the eye is exceedingly prominent, 

 and at it the temple bends rectangularly. In dimorpha the eye is 

 still prominent and large, but the angle is blunter. In heteracanthus 

 the whole temple from the base of the occipital band to the inser- 

 tion of the first antennal joint is evenly rounded, and the eye has 

 become elongate and comparatively inconspicuous. In hexakon, as 

 noted in the description, there is a curious short process from the 

 under side of the eye. We imagine this feature is less pronounced 

 in mirabilis. We cannot certainly say it is present in dimorpha, and 

 it seems to be wholly absent in heteracanthus. 



The gradual disappearance of this projection and the diminution 

 in the prominence of the eye are, we believe, to be correlated as 

 compensatory for the rise of the heterocerous condition. The $ 

 antennae are, we believe, accessory organs of copulation. In the 

 more primitive forms the female is held somehow between the 

 warped back antennae and the angle of the eye. Later the grasp is 

 made securer by the development of an appendage to the 3rd joint. 

 Finally a completely fast lock is established by (1) the processes from 

 joints 1 and 3, and (2) joints I and 2. There is only one weak spot in 

 this lock, viz. where the appendages touch one another, but if the 

 whole antennae is pressed back, as it doubtless is, against the head, 

 there can be no possibility of escape. 



The significant point in the development of these insects is 

 apparently when an appendage is produced on the 3rd antennal 



