336 THE AMERICAN NATURALIST.  [Vou. XXXII. 
closely with it, the only differences being due to a slight reduc- 
tion in the number of the veins ; the radial sector is reduced 
to a two-branched condition, the media is only three-branched, 
and only one of the anal veins is well preserved. 
Although it is an easy matter to determine the homologies 
of the wing-veins in a generalized form like Rhyphus, it would 
be exceedingly difficult, if not impossible, to do this in the 
case of some of the more specialized forms if they alone were 
studied. But when a carefully selected series of forms is 
examined the difficulties vanish. 
We wish now to call attention to such a’series for the double 
purpose of demonstrating the homologies of the wing-veins in 
the more specialized forms and of showing the value in taxonomic 
work of the characters presented by the wings. 
It should be borne in mind that the different parts of the 
wing may be modified more or less independently. Although 
the wing acts as a whole as an organ of flight, any change in 
the habit of flight is likely to result in a greater modification 
of some one part than of others. Thus we may find that 
in one line of deScent a certain part is greatly modified and 
another part remains but slightly changed from the primitive 
type ; while in another line of descent the opposite may be 
the case. It is necessary, therefore, in discussing the changes 
that have taken place in the venation of the wings to treat the , 
different veins separately. We will, however, refer to only a 
few of the more important of these changes, as a series of 
figures illustrating the homologies of the wing-veins of each of 
the families of this order has already been published by one 
of us: ; 
The reduction of the radial sector. — In a few genera of flies 
the radius retains the primitive, five-branched condition ; of 
these the genus Protoplasa of the Tipulide is a good exam- 
ple. But usually the number of the branches of this vein is 
reduced by a coalescence of some of the branches of the radial 
sector. Thus in many families the radial sector is three- 
branched, in others it is only two-branched, and in the gall- 
1 Comstock, Manual for the Study of Insects, pp. 413-489. 
2 Loc. cit., Fig. 504. 
