As many readers know, a group of us, known as the Parasitic Hymenoptera Advanced Research Team (PHART), have embarked on a project to key the New World genera of Braconidae. One problem that we have had to overcome is to agree on wing nomenclature. Most of us are settled on the fact that we will use some version of the Comstock-Needham system.
In the paragraphs that follow I will outline the Comstock-Needham system as it has been employed, for example, in the "Hymenoptera Of The World". Following this I will introduce some modifications that we are thinking of incorporating. I would like anyone who has criticisms or suggestions for improving the system to get in touch with me as soon as possible.
The Comstock-Needham system recognizes eight major longitudinal veins, abbreviated by capital letters. Starting from the anterior margin of the wing they are: costa (C), subcosta(Sc), radius (R), medius (M), cubitus (Cu) and 3 anal (1A, 2A, 3A). In addition, there may be a short jugal (J) vein in some Symphyta.
When a vein is branched the most anterior branch is given the subscript 1 and the more posterior branches the subscripts 2, 3.... This is done for all veins except R. When R branches the most anterior branch is called R1 but the second branch is called the radial sector (RS). [A note to readers: due to the present inability of html to represen subscripts (except as an image), I have been forced to omit them from the tex of Sharkey's article. nfj]
A vein may have several segments or abscissae. They are delimited by the intersection of other veins, usually crossveins. Thus a vein that is intersected by two other veins has three abscissae, numbered consecutively from the base to the apex of the wing, e.g., when Cu has three abscissae, they are 1Cu (the basal portion of Cu), 2Cu and 3Cu (the apical portion of Cu). Vein abscissae may vary interspecifically, therefore 3Cu of one species is not necessarily homologous with 3Cu of another.
Crossveins, indicated by lower case letters, take the name of the veins they connect, with the anterior vein given first. Thus, a crossvein that connects R with M is r-m. If there are several r-m crossveins they take numerical values as well e.g. 1r-m, 2r-m, etc. If a crossvein joins two branches of the same vein the crossvein takes the name of the major longitudinal vein, e.g. a crossvein between R1 and Rs is called r. This simplification is possible due to the rarity of this type of crossvein.
Two veins may fuse for part or all of their length, appearing as one vein. The resulting vein takes the name of both component veins joined by a plus (+) sign. For example, Rs and M are often fused for portions of their lengths. The fused portion is called Rs+M. Veins may fuse end to end so that it is impossible to know exactly where the first one ends and the second begins. In these cases the composite veins are joined with an ampersand (&). For example, in all Ichneumonidae and many Braconidae the first abscissa of vein RS+M is often completely lost and in these cases veins 1RS and 1M cannot be distinguished from one another: the composite vein is therefore termed RS&M.
Wing cells, abbreviated with capital letters, take the name of the vein lying anterior to them. If several fused veins form the anterior boundary of a cell, the cell takes the name of the vein that is theoretically most posterior. Thus, the cell posterior to C+Sc+R is the radial cell (R). If more than one cell is directly behind a vein, the cells are numbered consecutively from the base of the wing, e.g. three medial cells would be 1M, 2M, and 3M.
New wing veins may arise in certain lineages. Such veins are either given a new name or are named (often misleadingly) after the vein to which they are most similar in position. I follow the latter approach for convenience but distinguish the new vein from its namesake by an apostrophe ('). Thus veins followed with an apostrophe (') are not (or probably not) homologous with the same vein lacking an apostrophe (').
One problem inherent in the system described above is that although veins, crossveins and vein
branches are potentially homologous across the Hymenoptera, the abscissae are not. This is because the
abscissae of a particular vein are named based on the position of veins and
crossveins that intersect both anteriorly and posteriorly with the vein in question. To
see how this can be a problem refer to Figure 1a 
(upper figure, Helcon
fulvipes). Note that there are three abscissae of vein CU. Now refer to the fore wing of
Helcon pedalis(Fig. 1b). Here 1cu-a is positioned towards
the apex of the wing such that it intersects 1CU instead of being interstitial with M (I have exaggerated the
position for ease of illustration). In this case there is an extra abscissa of CU and none of the abscissae
of CU of the two wings (Figs 1a and 1b) are homologous.
To solve this problem the abscissa are numbered based only on intersection with veins or crossveins
that meet anteriorly. When crossveins meet the longitudinal vein in question from the posterior side a letter
is added to the name of the vein. For example the abscissae of vein Cu of Helcon fulvipes
are named according to this system in Figures 1c 
and those of
Helcon pedalis are named in Figure d. In comparing Fig. 1d)
has two components to abscissa 1 of vein CU whereas that of
Helcon fulvipes(Fig. 1c) has only one component of abscissa
1CU. Nonetheless abscissa 1CU of Figure 1c is easily homologized with
1CU of Figure 1d .
and b)
and of Helcon
fulvipes (Fig. 3a and b. These illustrate
some fine points and conventions of the system which are discussed below.
a) When two longitudinal veins meet they are presumed to merge unless there is evidence to the
contrary. For example veins 1RS and 1M meet to become RS+M. When two longitudinal veins meet they
do not adopt the abscissae letters and numbers of their constituent veins, rather they are numbered
independently. So in the case of the xiphydriid fore wing (Fig. 2a) the correct
name is RS+M not 1RS+1M. In Helcon fulvipes(Fig. 3a) the RS+M vein
is intersected from below by crossvein m-cu therefore there are two sections to RS+M, i.e., (RS+M)a and
(RS+M)b. The brackets ( ) are used to imply that the sections a and b refer to the composite vein RS+M
and not to the M component of RS+M which might be inferred if the vein was named RS+Ma.
Notice that (by convention) C+SC both terminate at or before the stigma of the fore wing or, in the case of the hind wing, at or before the point where R splits into R1 and RS.
When two fused longitudinal veins diverge, the new abscissae of the constituent veins take on an extra numerical value. Thus in the case of the xiphydriid (Fig. 2a), vein 1RS fuses with M to become RS+M and when they separate they give rise to 2RS and 2M (in this case 2M is divided into 2Ma and 2Mb).
I would appreciate your comments.
