■Jl'OLANDALES. 



187 



About. 15 somewhat (laboUuto, tlistallv forked, 

 and aiiastoinosiiiii; x'diiis radial.o from tho ossoii- 

 trial part of tlu) fruit, wliich is lartjo and spjicrical 

 and is sitiiat.od at tJui top of poiluuclo. No 

 l)rlmarics aro dilforout.iatod, but a vo'm runs to 

 tlio tip of e-M-h margimil lobe. Within tlio 

 meshes a system of still finer anastomosing 

 v(uul(!ts forms a four or five sided indistinct 

 areolation. The essiMitaal j)art of tlie fiuiti is 

 more or less globular, 5 to 8 millim*^ters in diam- 

 eter, and is adnate to the in\-oiucrr) at its l)aso 

 at the top of the pe(hincle. 



This species represents an undeseribed type. 

 The specimen shown in Plate X\T1, figure 4, 

 suggests a ligneous scale compara])le witli sonu! 

 conif(irous scale, but in reality it is an nntliick- 

 ened ^^'ing. These specimens, wliich are not 

 rai'e, have beoii compai'od with all the existing 

 famiUcs wliich have winged fruits tliat are 

 knowai to nw. They are more nearly compar- 

 able with certain existmg mendx^rs of tlu; 

 AnKHitiferie and are <ispecially suggestive of 

 Engelliai-dtia, which is repnisent<'d in tiie Wil- 

 cox flora liy perfectly characteristic wuiged 

 fruits as well as leaves. The essential part of 

 the fruit appears to be identical in both genera, 

 but in Paraengolhardtia the involucre is entii-e 

 and has only faint indications of the lobation 

 characteristic of EngeHiardtia, besides it lacks 

 the differentiated venation of that genus. It is 

 easy to understand that with the progressive 

 elongation of the incipient lobes of Paraengel- 

 hardtia, liecessaiy to a better dissemmation of 

 these fruits, the main vascular bmidle to the 

 tip of each lobe would become stouter and be 

 gradually transformed into midribs. This hhits 

 at the genesis of the Engelhardtia type of 

 fruit from ancestral fonns with small bracts 

 like the bracts of Juglans or Hicoria, which he- 

 came m the course of time concresceiit and sid)- 

 secpiontly deepl}' trUo])ate. 



Occurrence. — Lagrange formation (in beds of 

 Wilcox age), Puiyear, Henry County, Tenn. 

 (collected by E. W. Berry). 



Collection. — U. S. National Museum. 



Genus HICORIA Rafinesque. 

 Hicoria antiquorum (Newberry) Knowlton. 



Carya annquorun. Newberry, New York Lye. Nat. Hist. 



Annals, vol. 9, p. 72, 1868. 

 Newberry, Illustrations of Crotaoeons anrl Tertiary 



plants, pi. 23, figa. 1-4, 187S. 

 Newberry, U. S. Geol. Survey Mou. 35, pi. 31, figs. 



1-4, 1898. 



Carya anliquorum.. Lesquereux, IJ. S. Gciol. and (Jeog. 

 .Survey, Ten'. Ann. Ropt. for 1871, p. 29-1, 1872. 

 Lesquereux, idem for 1S72, ]>. ■102, 1873. 

 Ije.sfjuereux. The Tertiary flora, p. 289, pi. 57, figs. 

 1-5; pi. 58, fig. 2, 1878. 

 Carya antiqua. Lesquereux, l'. S. Nal. Mas. Proc, vol. 



II. p. 25, 1888. 

 Ifirorta anliquorum ('Xewl)erry). Knowllun, U. S. Geol. 

 Survey Hull. 152, p. J 17. lSi)S. 



De.sfTiption. — Lesquereux's description, pub- 

 lished in 1S78, is as follows: 



Leaflets large, broadly oval ur ovale-laneeolat*, acur ;- 

 naU', rounded or broadly euneate to the petiole; base in- 

 equilateral; l)orders minutely denticulate; lateral nervo.s 

 close, parallel, .•<iMiple, rin've<l in ascending toward the 

 borders. 



The leaflets of this species are very large, except those 

 of the lowest pair, whose size is, a.s in .species of .luglans, 

 generally diminutive. * * * • The substance of these 

 leaflets is subcoriaceous and rigid, the surface generally 

 polished, though deeply cut by numerous lateral nerves 

 and nervilles; the borders, erenulate or denticulate, be- 

 come entire toward the more or less inequilateral base, of 

 which one .side is generally rounded, the otlier straight. 

 The petiole of the lateral leaflets is short; that of the ter- 

 minal oives longer. * * * The lateral nerves are close, 

 twenty to twenty-five pairs in the large leaflets, under a 

 broad angle of divergence, 50° to 60°, mostly simple, 

 closely following the borders in simple bows, connected 

 with the teeth by minute short nervilles. * * * The 

 generic relation of the species can not be definitely con- 

 sidered as long as the fruit is not kno\\'n. As the North 

 American species of Juglans and Carya can be used only as 

 points of comparison by the characters of theii- leaves, these 

 characters, especially the generally simple secondary 

 nerves in our species of Juglans more generally di\'ided in 

 those of Carya, seem to refer this fine species, whose lateral 

 nerves are not at all divided, to the first genus. The size 

 of the leaflets, however, has more likeness to those of 

 Carya alha. tliough all the fossil leaflets of Carya published 

 by European authors are naiTow and linear, and also the 

 branching of the nerves is quite as distinct in Juglans 

 rupesfris as in the species of Carya. 



Th(^ foregoing quotation is somewhat 

 abridged from Lesquereux's description of 

 this species, which is found in considcra])le 

 abundance in the W^ilcox at ( -amjjbeU's quarry, 

 La. The form is widely distributed in the 

 basal and Fort Union Eocene of the Rocky 

 Motmtam province, extending northward into 

 the Northwest Territory and British Columbia, 

 accor<ling to PenhaUow, whose records 1 htive 

 not quoted in th(> synonymy because of the 

 extreme unreliability of most of his unillus- 

 trated determinations. 



W^ith regard to the generic reference of these 

 leaves to Hicoria, 1 believe it to be entirely 

 unsatisfactory but have not ventured on a 

 change, since that would obscure the geologic 



