74 AUT. BOX. 



and Chrysophylum, — Sclerozus means hard 

 branches, they are brown rugose, leaves petio- 

 late cuneate acute or obtuse, reticulate above, 

 glaucous silky hairy benpath : flowers fascicu- 

 late on long peduncles, calix camp, rufous, 5 

 lobes obt. unequal, corolla rather longer. My 

 specimens are inerme, from Florida. 



503, BuMELiA undulata Raf. new flora 545 

 — Cuba and South Florida, 



504, Bum. arachnoidea R. n, ff. 546 — Ar- 

 kanzas. t 



505, Bum. rufa Raf. ramis fusco rufis villo- 

 sis, fol. subsessil. ellipt. obt. undul. margine re- 

 vol. sapra glaiicescens subtus rufis villosis reti- 

 culatis ; flor. glomeratis densis rufis — Florida, 

 disc, by Warie, I made it a var. of B. lanugi- 

 nosa in new flora, but it appears peculiar, t 



506, Bum 1 serrulata R. n. fl. 548 — Alaba- 

 ma, a doubtful shrub, since flowers lacking: 

 there are many such in- Florida and Alabama, 

 even when flowers known, as the 3 next till 509. 



507, Bum ? denticulata R. n. fl. 547, cer- 

 tainly no Bumelia, I was deceived by Collins 

 label ; he and I have destroyed all the flowers 

 except 2 to analyze this singular tree, which 1 

 compared with Nyssa dentictdata and SchU 

 zandra, both akin in leaves, but not in flowers. 

 It is nearer to my G. Lomilis, but I have fixt it 

 protem as follows. 



STREBLINA Raf. (ambiguity) dioical, fl. 

 masc . . . fem. calix 5part. lac. ineq. subrot. ob- 

 tusis nndnl. petalis 5 elHpt. obt. undul. subeq. 

 calix brevior, pistilum minimum in apex pedun- 

 culo immerso,stigma obt. sessil, fruct . . . drupa 

 Isp ? — Thus it diflfers from Nyssa by petals, un- 

 equal calix &c, from ScHizandra hy many pe- 

 culiarities, and yet it may be of same family or 



