234 MR. A. W. HILL—REVISION OF 
known and more intelligently explored it will no doubt be possible to get a clearer idea 
of the relationships of the various forms to one another. 
At present the ripe fruits are known from only a few plants, and without doubt they 
form a very important basis for classification. Four species are defined in this paper. 
(1) N. pichinchensis.— The well-marked large form, with conspicuous mauve flowers 
about 3 cm. in length, of which we have no ripe fruits. It is of this plant that Jameson 
says in a note at Kew :—“ S. pichinchensis occurs only between the elevation of 15,000- 
16,000 feet. It forms a sort of dense turf, and, from the beauty of the flowers, enlivened 
in no small degree the gloomy aspect of the surrounding scenery,—the Chuchiraga 
insignis is the only shrub reaching the same level. We likewise met with several 
Composite, such as Culcitium and Espeletia grandiflora.” It occurs on Pichincha and 
on Chimborazo * more sparingly (Jameson, 1859, ann.) A well-marked varietal form of 
the true W. pichinchensis, growing mainly or entirely on Chimborazo, differs from the 
true form in having the smaller leaf-lamina much more dissected, with the lobes longer 
and narrower and the corolla only about 2 cm. in length. These forms are probably 
identical with the form mentioned by Hieronymus t from Chimborazo. 
(2) N. phyllanthos.—Closely allied to the true N. pichinchensis are the forms which 
I believe to be the Sida phyllanthos, Cav.}, and S. saxifraga, H. & B.$. They are 
represented in the herbaria by no. 2257, Bonpland, from Antisana (Herb. Mus. Paris. 
and Herb. Berol), and by some specimens of Hartweg, 917 |, and of Whymper from 
Antisana and Pichincha. Their distinguishing characteristics are that the petioles and 
laminz are hairy on both sides, more or less densely, and the stipules are adnate only 
to the middle of the petioles, which are about 1°5 cm. long. The lamina is trifid, each 
segment being trilobed with obovate-obtuse lobes, entire or again more or less 3-lobed ; 
and in these points the plants accord very well with the descriptions of Cavanilles. 
It is from Hartweg, 917, that the only ripe fruits of this group of forms are known. 
Another form very closely allied to the preceding may very possibly be only a varietal 
form. It is, however, much more like a dwarf form of N. pichinchensis. The leaves are 
glabrous and varnished on the under surface and the stipules are adnate almost to the 
lamina. The lamina, which is really trifid, tends to become palmatifid by the development 
of the two external lobes, and the median lobe tends to be 5-lobed as in JN. pichinchensis. 
The type-specimen of this form is Jameson’s no. 32, in which the leaves are closely packed 
in small rosettes only about 3 cm. in diameter. This specimen is labelled Sida phyllanthos 
both at Kew and at Cambridge and comes from Antisana. With it agree the forms 
labelled Hartweg, 917, at Kew, a Cambridge specimen from the herbarium of Dr. Leman», 
named by Bentham, and also a plant at the British Museum collected by Whymper on 
Antisana €. 
* According to Whymper, ‘Travels among the Great Andes of Ecuador,’ p. 199, M. pichinchensis was found on 
Antisana at an elevation of from 15,000-16,000 feet. 
T Hieron. in Engl. Bot. Jahrb. xx. Beibl. no. 49, p. 43. 
+ Cav. Diss. v. 276, t. 127. f. 4. $ Humb. et Bonpl. Pl. Æq. ii. p. 116 adnot. 
| Hartweg, 917, Colombia, in Herb. Dr. Lindley in Herb. Univ. Cantab., Herb. Kew., and Herb. Mus. Brit. 
«| v. Whymper, l. c. p. 199. 
