NO. 2146. SOME AMERICAN FOSSIL INSECTS— COCKERELL. 99 



obtuse at end, scarcely or not open. The following measurements 

 are in microns: 



Cross vein connecting Ri witli Sc. basad of (vertical) level of base of 



first fork. 



400 



Base of discal cell to base of first apical fork (separation of R^ from Ra). 900 



Ra from its separation from Rs to end 1 7(50 



R4 from its separation from R5 to end 2, 208 



Base of second apical fork basad of level of base of third fork 560 



Base of fourth apical fork basad of level of base of third fork 400 



Ml from its separation from M2 to end 1 §40 



Ms from its separation from M4 to end 1, 700 



Base of fourth apical fork apicad of level of base of fifth fork 1, 120 



Cui from its separation from Cuz to end 2, 320 



End of anal cell apicad of level of base of fifth fork 4S0 



In amber from the Eutau formation (Upper Cretaceous; Em- 

 scherien), Coffee Bluff, Hardin County, Tennessee (Bruce Wade). 

 The specimen was very kindly forwarded by 

 Prof. Edward W. Berry. It is of extraor- 

 dinary interest, as being the first insect to -c:^;^__^ r— 

 be described from American amber. If "^---^J^ 



any considerable insect fauna can be found 

 in Cretaceous amber, it will undoubtedly 

 throw much light on many obscure prob- 

 lems connected with the origin of the mod- ^^'''- 6.— dolgphili-.s pkaemis- 



j. .,. T ?.-,■,.-,-., ^^'S- Anterior wixG. 



ern families and genera. Dolophilus Mc- 



Lachlan (Philopotamidae) has four species in Prussian amber, of 

 Oligocene age. The present insect is too close to such forms as 

 D. aequaUs Hagen to be generically separated, so far as the visible 

 characters show. Ulmer remarks that as regards venation Dolo- 

 philus is extremely like Phijlocentropus (Polycentropidae), and I 

 was indeed in doubt whether to refer the present species to the latter 

 genus. The discoidal cell in Pkylocentropus is shorter, and on the 

 whole our species seems rather to belong with DolopUlus. Accord- 

 ing to Ulmer's phjdogenetic scheme, the Philopotamidae are more 

 primitive than the Polycentropidae. It is also noteworthy that of 

 the four living species of Dolouhilus three are European and one is 

 Australian. 



Thus we may say that our Cretaceous fossil is of a very modern 

 type, but it would be more correct to put the matter another way and 

 say that the living Dolophilm is a remnant of a very ancient group. 

 It is, in fact, closely related to the Xecrotauliidae of the European 

 Lias. 



Type.—C&t. Xo. 62001, U.S.N.M. 



