126 
ROYAL HORTICULTURAL SOCIETY. 
Passiflora tomentosa, var. p. moUismna, Triana and Planchon, 
in Ann. Sc. Nat., Ser. 5 (187.3), xvii., 131. 
IS'ew Granada. 
Messrs. Triana and Planchon state that they find no difference 
between the Peruvian Tacsonia tomentosa and the New 
Granadan T. mollissima, except that the flower- tube is 
slightly pubescent in T. tomentosa^ while it is quite glabrous 
in T. mollissima In this opinion I cannot concur. The 
leaf-lobes of T. mollissima^ as I understand it, are ovate- 
lanceolate or lanceolate, instead of broadly ovate ; they are 
downy on both surfaces, instead of being nearly glabrous 
above, as in tomentosa. The involucre of T. mollissima 
often splits on one side like a spathe, and is composed of three 
ovate -lanceolate segments, which are united one to the other 
at different levels, and which have everted edges. In T. 
tomentosa^ Lam., the involucre is regularly tubular, the con- 
stituent bracts are united two-thirds of the way up, and the 
free portions are broadly ovate-acute. 
7. T. mixta, Juss., in Ann. Mus., vi., 394; Mast, in Flor. 
BrasiL, I.e., 541. 
Andes of New Granada, Ecuador, Peru, and Bolivia. 
Var. 1. spECiosA, Mast., I.e., 541, tab. 128, fig. 1. 
T. spediosa, H.B.K., Nov. Gen. et Sp., ii., 43. 
T. tomentosa, Juss., var. speciosa, Mast., Gard. Chron,, 
1870, 955 (description and comments); Florist 
and Pomologist, 1871, 169. 
Var. 2. QuiTENSis, Mast., I.e., 542 
Tacsonia quitensis, Benth. Plant. Hartweg., p. 183; 
Mast, in Gard. Chron., 1869, 388. 
T. tomentosa, JjEim., var. quitensis ; Triana and Planch., 
Ann. Sc. Nat., xvii., 1873, 131. 
Var. 3. EEiANTHA, Mast., I.e., 542. 
Tacsonia eriantjia, Benth. PI. Hartweg, p. 183; Hook. 
inBot. Mag., 5750. 
T. quitensis, var. eriantha ; Mast, ia Gard. Chron., 1869, 
388. 
The arrangement above given is that adopted in my monograph 
in the Plora Brasiliensis. The disposition proposed by MM, 
Triana and Planchon, I.e., is a little different, but considering 
how variable these plants are, how wide their geographical 
distribution (in some cases), and how great are the facilities 
for intercrossing, it would be impossible to obtain perfect 
