8 PKOCEEDINGS OF THE NATIONAL MUSEUM vol.66 



The reference to Lomatia follows the usage for the Florissant 

 fossils but actually the genera of Proteaceae can hardl}^ be separated 

 on the leaves. The living Lomatia ferruginea E. Brown and L. 

 tinctoria R. Brown have foliage of the general type of Grevillea 

 robusta A. Cunningham; while L. ohliqua R. Brown, L. dentata R. 

 Brown, and L. polymorpha R. Brown are entirely different. Were 

 they all fossilized, they certainly would not be regarded as con- 

 generic. I take this opportunity to note that the reference to fossil 

 leaves resembling Grevillea^ in American Museum Journal (vol. 16, 

 p. 449), has to do with Lomatia acutiloba. The editor (p. 447), un- 

 fortunately inferred that the species figured (Z. tripartita) was re- 

 ferred to, and this was later the occasion for a criticism from a South 

 African botanist who does not believe in North American Proteaceae. 



BANKSITES LINEATULUS, new species 



Plate 2. fig. 3 



Seed about 2.5 mm. long and 1.5 broad, with wing 5.8 mm. long 

 the base of which falls short of end of seed about 1.6 mm. ; the wing 

 is 3 mm. broad, with six or seven widely spaced delicate veins ; it is 

 obtuse and inequilateral, and wing and seed together measure 7 

 mm. in length. 



Green River Eocene, Roan Mountains, Colorado, at Station 2 of 

 1922 expedition, large excavation wuth tunnel at head of Salt Wash, 

 some distance below top of hill (Cockerell). 



H olotype. —C2it. No. 36860, U.S.N. M. 



This is extremely similar to B. lineatus Lesquereux from the Mio- 

 cene of Florissant but smaller and presumably a different species. 

 It differs little from the seeds of living Proteaceae; there is even 

 some suggestion of the projecting point near the upper end of the 

 seed which is seen in Banksia. In Banhsia integrifolia Linnaeus 

 filius, the wing goes less than half way down the side of the seed ; in 

 Banhsites lineatulus it goes much more than half way, but not 

 nearly so far as in Knightla excelsa Robert Brown, in which it goes 

 practically to the end. 



The fossil leaves described as Banhsites saportanus Velenovsky, 

 recorded from the Upper Cretaceous of Marthas Vineyard, are 

 much more like Kniglitia than Banksia. 



Family HAMAMELIDACEAE 



LIQUIDAMBAR CALLARCHE, new species 1 



Plate 1, fig. 6 ; plate 2, fig. 5 



Leaves similar in size and appearance to those of the living L. 

 styraciflua Linnaeus, five-lobed, the lobes without accessory lobules, 

 the basal margin (dentate in L. styraciflua) entire, sides of median 



