74 PROCEEDINGS OF THE AMERICAN ACADEMY 



Bquamis termi-subiilatis imae basi corollas insertis antheris cum columna 

 breviore (sed manifesta) adaequantihus. — M. parvijioruin, Cbapm. 

 Fl. 367, non R. Br.— Pine Key, S. Florida, Blodgett. 



GONOLOBUS and LACHNOS TOiAIA. Later autbors generally 

 appear to liave taken tbeir idea of Gouolohus from tbose numerous 

 tropical speoies wbicb ditler considerably from tbe typical ones of tbe 

 Soutliern Atlantic States. Tbe latter have not the coloied-reticulated 

 corolla of most of tbe tropical species, nor tbe broad and rounded 

 lobes of many of them, nor tbe strongly salient-angled stigma, which 

 has been made a part of tbe generic character, nor is the corona borne 

 on tbe corolla, but either distinctly on tbe short stamineal column or at 

 its very base at tbe junction with tbe corolla. Laclmosfoma^ 11 BK., 

 was founded on a single and peculiar species, with " corolla subbypo- 

 crateriformis," or at least with tube as long as the limb, and with 

 "corona pentapbylla, fauci coroUte inserta, foHolis cuneatis carnosis 

 apice lunato-bilobis," etc. Noting that tbe column of (llatnents is 

 aduate to tbe tube of tbe corolla (as it truly is for tbe whole length), 

 Kunth, indeed, adds the query : '' An igitur corona summo tubo fila- 

 mentorum nee fauci coroUae inserta?" But the flowers I have exam- 

 ined, from Fendler's no. 1050, show that the first view was tbe correct 

 one. On separating the tube of the corolla from that of the filaments, 

 almost without lesion, the coronal appendages are left on tlie former, 

 to which they manifestly belong. There are, moreover, vestiges of a 

 stamineal corona, in tbe form of minute scale-like processes, one under 

 tbe base of each anther. Decaisne adds three other species to the 

 genus, and characterizes it by the " corona staminea gynostegio plus 

 minusve adnata, pentapbylla." It will be allowed that, if tbe genus 

 can stand upon tbe 5-pbyllous corona, tbe mere insertion will be 

 unimportant. But Mr. Bentbam, enlarging the genus much more by 

 adding Decaisne's Ibutia and Chthamalia, includes species with cup- 

 shaped and wiih annular corona, thus resting LacJmostoma mainly ou 

 tbe insertion of this oi'gan, which is variable from species to species, 

 and really not different fi'om that of typical species of Gonolobus. 

 And our only reticulate-petaled species of the latter bears the annular 

 corona distinctly on the column, close to its base, but wholly separate 

 from the corolla. It thus appears, on tbe whole, that LacJmostoma 

 should be restricted to the original L. tigrinum ; and that CJdhamalia 

 (as the subjoined details show) should form only a subgenus of Gono- 

 lohus. Those reticulate-flowered tropical species (of which we have 

 one, not very typical, representative in Texas), with all its variations 

 in the gyuostemium and corona, may form another section, which may 



