HYMENOPTERA SAW-FLIES, ETC. 169 



of the wing, and have the appearance of branching out of the sub- 

 costa. One leaves it just before its junction with the costa (at a 

 point about h), and, bending first down and then a little up, finally 

 joins the costa (on the margin of the wing) at its apex k. This is the 

 radius. The other called the cubitus leaving the sub-costa earlier 

 (i.e., nearer its base) at /, bisects, roughly, the area between the 

 radius and the apical portion of the medius, and so proceeds not 

 always in so straight a line as the figure shows to the point y on the 

 margin. 



" Thus we have in all five main and two subsidiary longitudinal 

 nervures or 'veins,' and these divide the wing into longitudinal 

 areas or 'fields,' as follows. 



" Between the costa and the sub-costa is contained the intercostal 

 field. Between the brachius and the humerus is the humeral field, 

 familiar to all students of saw-flies under the name of the ' lanceolate 

 cell.' Then between the sub-costa and the medius lies the medial 

 field, and between the medius and the brachius the brachial field. 

 (A certain similarity in form and size will probably have been noticed 

 by the reader between the intercostal and humeral fields and the 

 medial and brachial fields respectively. This will help him perhaps 

 in forming a mental picture of the neuration as a whole.) 



" Next we have the radial area lying above the radius, and the 

 cubital area below the radius and above the cubitus. Between the 

 cubitus and the apical half of the medius comes a continuation of the 

 medial area, which, if regarded as distinct from it, may be called the 

 discoidal field. That part of the wing lying below and as it were 

 outside of the neuration system i.e., that which is bounded in- 

 feriorly by the actual inferior margin of the wing and superiorly by 

 the humerus as to its basal half, and by the medius as to its apical 

 portion is called the anal field. 



" Owing to the disappearance of the brachius without reaching the 

 margin, there is no complete longitudinal line of division between the 

 brachial and anal areas. But for practical purposes the transverse 

 nerve, / x, may be regarded as separating them. 



" We come now to the transverse elements of the neuration 

 Konow's ' Nerven ' sensu restrido. 



" Often (though not in my figure) the radial area is crossed by one, 

 very rarely by more than one. Then the radial area is said to be 

 'divided,' or, as some authors express it, 'there are two radial cells.' 



