DESCKU'TION OF THE SrECIES. 283 



polygonal in form, quite different from those next al)nve them in size. This 

 figure does not show any of the original margin oi the leaf or of the pet- 

 iole, althougli the outline and dimensions of the leaf are evidently retained 

 nearly unchanged. 



This curious leaf, in shape and nervation, remind.s one of the leaves 

 called Protorhipis by various authors, under the supjiosition that they are 

 the prothalli of ferns. It may well Ijo questioned, however, whether some 

 of these are not ancient forms of dicotyledons. The nervation of the Poto- 

 mac plant is something like that of a fern, and is decidedly archaic in type. 

 The form is similar to that of some Hedcra, as H. prlmordiaUs Saporta, but 

 in Hedcra the midrib and primary nerves are much stronger than the sec- 

 ondary ones. In the midrib splitting up into numerous primary nerves of 

 nearly equal size, in the dichotomous forking of the latter, and in the very 

 thick petiole we have a strong resemblance to some of the Protcacece, as 

 Protca cordata Thunb. In shape it is a good deal like Begonia hydrocotyli- 

 folia. It is also similar in shape to Popidtis arctica Ileer.' This species 

 may be taken as the type of the subgenus Rotundutmn of the forms of Pro- 

 ie(rp1iyUiim. 



Pl'OTEyEPIIYLLUM ORIilCULARE, Sp. nOV. 

 Pl.atc CXXXIX, Fig. 4. 



Leaf small, orl)icular in slia[)e, slighth' heart-shaped at ba.se; primary 

 nerves all nearly of equal thickness, diverging slowly, and radiating from 

 the base of the leaf, sending off obliquely very slender branches, which 

 anastomose to form irregular comparatively large meshes ; ultimate nerves 

 forming oblong pentagonal or hexagonal meshes. 



Locality : Fredericksburg ; only one specimen found. 



This plant also belongs to the rotundate section of Protecpphi/Uum. 

 It more than the preceding species resembles Hedera, but the radiate 

 arrangement of the slender primary nerves does not agree with that genus. 

 It may also be compared with Protea cordata Thunb. The nervation of 

 this plant is much like that of Banksites Saportanus, Velenovsky (see Die 



' Flor. Fo.ss. Arctica, vol. 7, I'l. I.XVII, Fig. 2. 



