XIX. 



form a now organ, or one that mighl easily be mistaken 

 for some other part. In the fourth species, rupicola, the 

 costa reverts to its normal position. The four species are, 

 however, most closely allied in all other directions, and thm 

 it is clear, the sacculus being olten quite distinct, that the 

 valva is a composite organ, composed of three separable 

 parts. 



The term harpe is always rather indefinite, and in the 

 Tortricida: would generally be synonymous with (he costa, 

 we therefore do not employ it. 



In the Geometrido- the term furca was employed in 

 those cases were the sacculi were separated from the 

 valvula 1 , and united centrally, so as to form one organ. 

 The term, it appears, had already been employed to 

 represent other organs in the Lyccenidcc. In the present 

 volume we therefore describe such cases by saying that the 

 sacculi are centrally united. 



The eighth segment, as we should expect from the 

 fact that the genitalia are not withdrawn behind it, is 

 remarkably devoid of ornamentation, though occasionally 

 coremata exist. 



We have been critieised for dividing the parts 

 and organs by the expressions external and internal. 

 We would point out that some parts develop within the 

 abdomen, and only become external at the end of this 

 process, at which time they also become fused to the 

 tegumen. These parts we call the internal parts, but we 

 have to admit that the valvae should, strictly speaking, be 

 reckoned with them. The penis, as we nave said, is very 

 characteristic of the family. The aedoeagus is generally 

 much curved, and hinged to the central, shield-shaped, 

 juxta; thus the elaborate arrangement surrounding the 

 penis, so prominent in the Gcometrida, is rendered 

 unnecessary. The eedceagus is as a rule simple, sometimes 

 it is dentate, very rarely branched. The vesica is not only 



