152 DE- E- J- TILLYARD ON THE 
(2) M has become fused basally for a short distance with Rs, the basal 
stem of the two veins becoming obsolescent, but actually showing a weak 
connection with M only, while the lost origin of Es is shown by a \ery 
definite stump projecting from Ri close to the base. In some recent 
Mayflies, however, Rs remains separate from if, and'is attached tangentially 
to Ri just- beneath km. Both conditions may be regarded as specialisations 
involving the abortion of the original basal piece of Rs. 
(3) The basal portions of M and Gu, before the first forkings of these 
veins, run close together as in Protereisma, but both of them turn upwards 
almost transversely across the wing. At the upper end of this bent portion 
is the formation which I have called the cubito-median Y-vein, so greatly 
Text -FIG. 4. 
Basal portion o| fore-wiug of Aiiieletus oniatus (Eaton), Recent, considerably enlarged 
to show details of venation, k, stump of Rs. Above, costal vein and humeral 
brace [hm) in Mesephemera ceUtilosa (Hagen), Upper Jurassic of Solenhofen. 
For lettering, see p. 162. 
reduced from its original conditions (as seen in Proteresima) that it can 
only just be made out in text-fig. 4. At this point, Cu forks, and its two 
branches turn at right angles to its former course and run outwards about 
parallel to the posterior border of the wing. Cuj remains, as ia the 
fossil, a strongly-curved vein concave to the posterior border; but the 
shortening of its length has brought with it a definite increase of curvature, 
so that the weak sigmoid curve which this vein shows in Protereisma 
becomes replaced by a single strong arch, concave to the posterior border. 
Other recent Mayflies, such as Ephemera, have the original sigmoid curve 
intensified, but the vein itself greatly reduced in length, and the clavus 
almost completely obliterated. 
