10 SCUDDER ON THE DEVONIAN 



requires a very unusual development of the internomedian area, leaves the lower externo- 

 median field in its usual proportionate extent as compared to the upper field, and is 

 further supported by several considerations : chiefly by the probability that where 

 repetitions of structure are found — a mark of simplicity much more common among 

 ancient than among recent insects — they are far more apt to occur between repetitive 

 parts than between those which may not be so exactly compared. On the hypothesis 

 sustained above, this repetition occurs in the fields embraced between the two similarly 

 disposed sets of branches into which one vein is divided. On the other suggested (and 

 apparently the only alternative, for the open interspaces on either side of the lower branch 

 of the upper externomedian stem seem to fix that nervule unquestionably) the repetition 

 Avould be between the whole of one set of branches of this vein, and one portion only of 

 the two of which the other set of that vein is composed. Other arguments may be 

 advanced from the character both of the nervules and of the cells formed by them and the 

 cross veins, which differ slightly from those in the field next above, a difference greater both in 

 extent and in nature than that existing between what we have considered the upper and 

 the lower externomedian fields. Further than this, the slight change of direction in the 

 course of the outer margin, resulting in a slight emargination of this border of the wing, 

 although apparently not found at all in living Ephemeridae, would be far more likely to 

 occur, does far more frequently occur in other insects, between two adjoining areas than 

 in the middle or other part of one. 



Considering then the field under discussion as belonging to the internomedian area, we 

 must describe this as plainly of very unusual extent, and as filled as it never is in living 

 types with a large number of intercalary nervules. 



It may be remarked that none of the many intercalaries in this wing arise indepen- 

 dently, and that they are not more abundant at the extreme outer edge of the wing, as 

 is frequently the case in modern types. The former feature is the more noteworthj^ as 

 the independent origin of the intercalary veins in Ephemeridae would naturally be 

 taken as a mark of inferior organization ; and yet it does not occur in this oldest member 

 of the group, nor yet in the Jurassic species from Solenhofen, described on a previous 

 page ; in this last, hoAvever, the edge of the wing is more broken by intercalaries than 

 the parts removed from it. 



The length of the fragment preserved is 42 mm. and its greatest breadth, 25.5 mm. 



The points in which this insect presents the most striking differences from modern 

 types, and upon which we would establish the genus Platephemera, are : the very similar 

 instead of distinctive structure of the framoAvork of the two sets of branches of the 

 externomedian vein, and of the respective areas included between them ; the excessive 

 number of the intercalaries in the area included between the lower set of externomedian 

 branches, and their attachment (in the apical half of the Aving) to the upper of these 

 branches — from which the prcA'iously mentioned peculiar feature mainly depends ; the 

 simplicity of the lower branch of the upper externomedian stem in an unusually ramose 

 wing ; the unusual extent of the internomedian area and its rich supply of intercalaries ; 

 the density and polygonal form of the cells formed by the cross veins below the upper 

 externomedian vein ; the emargination of the outer border ; and finally the vast 

 dimensions of the wing. 



