E. A. COCKAYNK 81 



III the lobster, Honuirus vulgaris, described by Nicholls before the 

 Royal Society in 1730, this coiulition was also present. 



There was jjerfect halving in all respects both external and internal. 



Though perfect halving of external parts is common in insects, it is 

 very rarely met with in the genital apparatus. In Lepidoptera only a 

 single instance is on record, that described by Toyama in Bonihyx viori. 

 But we cannot suppose that there is any fundamental difference between 

 the halved gynandroniorphous insects and birds. The most simple 

 explanation of the very varied appearance met with in the sexual 

 apparatus in halved gj'nandromoi'phous insects is that there is a 

 tendency for the halves of the germinal epithelium and primitive 

 genital tract each to form a complete sexual apparatus, including two 

 gonads, by completion (if unpaii-ed median organs and doubling of 

 paired lateral organs. 



Sometimes the doubling does not affect the gonad, but only the 

 accessary glands or ducts or the external genitalia. 



Associated with this there is often a failure in development of part 

 of the sexual apparatus of both sides. 



Examples of this are the Melitaea didyma and the Saturnia spini 

 figured by Wenke, each with two testes on one side and two ovaries on 

 the other, with many of their respective accessory organs. In both 

 cases the glandulae accessoriae failed to develop. 



Examples where the doubling has only affected the lower part of 

 the genital tract are two of the three specimens of Amorpha populi 

 figured and described later (No.s. 6 and 8, Diagrams / and h). 



Sometimes one half shows a tendency to develop perfectly, with or 

 without doubling, whilst the other half fails to develop in its lower pai-t. 

 An example of this is afforded by the A. populi figured and described 

 later (No. 7, Diagram g). 



If the upper part of the tract, which forms the gonad, fails to develop 

 on the one side, as it does very freijuently, a primary somatic hermaph- 

 rodite is produced. The other half may produce a single set or a 

 double set of sex organs, and the doubling may affect the gonad. 

 Examples of this condition with doubling of the testis are specimens of 

 A. populi, such as the one I figure later; with doubling of the ovary, 

 an ^'1. populi and a Vanessa aidiupa figured by Wenke, and with a single 

 ovary, a Dryas paphia also figured by Wenke, Gerstacker's A. populi^ 

 and an ant, Leptothorax tuberum, described by Wheeler. 



On the side, which has failed to develop its gonad, all the other 

 sexual organs may be present, or only the lowest part of the internal 



Journ. of Gen. v (> 



