574 REVISIONAL NOTES ON AUSTRALIAN CARABIDiE, V., 



natural groups in which resemblances are found become small- 

 er, inheritance may be given more weight, so that, speaking 

 generally, we need seldom go beyond reversion to explain such 

 resemblances ; but, it is not always the case that heredity, or 

 atavism, is the true key to the riddle of similarity, even in 

 species that are not remote from one another. 



The difficulty of distinguishing between convergence and 

 reversion is often very great, and the interpretation of such 

 cases depends largely on the mental attitude of the inquirer. 

 The nearer the relationships between the species involved, the 

 greater becomes the perplexity as to whether doubtful in- 

 stances of likeness in structure should be assigned to conver- 

 gence, or to reversion. There is a third category in which 

 should be placed cases of reversion with convergence, as, when 

 two or more descendants, from the same stem, independently 

 converge towards such ancestral stem along the same lines of 

 reversion. It may be said this is merely reversion ; but, when 

 two descendants from one ancient type, after differentiation 

 and modification of more than one character, revert towards 

 their common ancestral form, each by a different character, 

 there is reversion but no convergence ; so when there is both 

 reversion and convergence, a distinctive term, such as rever- 

 sionary co7ivergence, becomes useful. 



The resemblance between the form of the lobes of the men- 

 tum and the inner lobe of the maxillae in the genus Helluo, 

 and in Gigadema longipenne, I believe to be due to reversion- 

 ary convergence. As I interpret the perplexing case of the 

 resemblances between the mentum and the inner lobe of the 

 maxillae in Helluo and G. longipejiJie, these resemblances are 

 due more to convergence than to reversion. The following 

 evidence may be offered in support of this view. The long, 

 pointed lobes of the mentum and the ordinarily shaped maxil- 

 lae of G. longipenne are exceptional in the genus Gigadema, 

 but are the ordinary form in Helluo (two species, not nearly 

 allied). Therefore we may feel justified in believing the pro- 

 totype of Helluo had these characters, which are also found in 



