22 Journal New York Entomological Society, tvoi. xxiv, 



1890, has already pointed this out." Walton gives no reason for this 

 statement, and Brongniart's work is not now accessible to me, so that 

 I do not know what evidence he has adduced in support of this con- 

 tention. If one will compare Fig. 8 (Plate II) of Stcnodictya with 

 that of the Mantid, Chocradodis, depicted in Fig. 9, however, it will 

 be apparent that the paranota of the Mantid's pronotum occupy 

 exactly the same location as the prothoracic paranota of the fossil 

 insect (Stcnodictya), and why they are not to be considered homo- 

 logous is not apparent. It is not evident that the paranota of the 

 fossil insect's prothorax are articulated or movable, and even though 

 they were, the case of the beetle Acrocinus cited above, would indi- 

 cate that pronotal paranota may be articulated even in existing forms. 



In reply to the argument that the precursors of wings must have 

 served some useful purpose while developing, one might state that 

 when we are dealing with an inherent stem tendency (as the develop- 

 ment of paranotal projections seems to be) it is unnecessary to postu- 

 late that the products of such a tendency must serve some useful pur- 

 pose — so long as the result is not detrimental to the organism, where- 

 upon natural selection would operate to check further development 

 along such lines. It is not beyond the realm of possibility, however, 

 that the paranota may have served a useful purpose (before becom- 

 ing adapted for gliding flight) in that they may have served to shield 

 the flanks and basal portions of the legs, as Grassi has pointed out. 



It is evident from the foregoing discussions, that the objections 

 which might be raised to the paranotal theory are not insurmountable. 

 On the other hand, there are certain insurmountable difficulties in the 

 way of accepting the tracheal gill theory, as were shown in points 

 (i), etc., although these difficulties do not apply in the case of the 

 paranotal theory. Furthermore, the paranotal theory is in full ac- 

 cord with all of the known facts — which is more than can be said for 

 the tracheal gill theory ! 



In making a choice between two rival theories, our selection 

 should not only be free from insurmountable objections (or should 

 not be incompatible with any of the known facts), but should also 

 have the positive quality of being in full accord with all of the known 

 facts. If we apply this criterion in selecting one of these theories, 

 the paranotal theory, being the only one which fulfils the conditions, 

 must therefore be chosen, and has been here adopted as a provisional 



