422 Hill .— The Genus Calt ha in the Southern Hemisphere . 
followed by later writers ; but it is convenient to relegate the Southern species 
to a separate section, for which the name Psychrophila may be maintained, 
on account of the peculiar development of the appendages of their leaves. 
Eleven species are recognized in the present paper, three being described 
for the first time. The new species, though collected many years back in 
the high Andes of Ecuador, Peru, and Bolivia, and in Tasmania, have 
apparently never received critical examination. The Andine plants have 
up till now been included under C. sagittata^ Cav., which is confined to the 
Magellanic region and the Falkland Islands, 1 and the new Tasmanian 
species has been placed under C. introloba from Australia, from which, how¬ 
ever, it is markedly distinct 
The characteristic feature of the plants belonging to the section Psychro¬ 
phila , as now understood, is the development of the auricles of the leaf 
laminae to form upturned or erect appendages, and this peculiarity is seen 
in its simplest form in C. sagittata and C. novae-zelandiae, Hook. f. In 
these two species the elongated auricles or lobes of the sagittate leaf are 
sharply folded up from the base, and the appendages thus formed either 
stand somewhat erect or lie partially over the face of the laminae with their 
under surfaces facing upwards. In C. andicola, Gay, and C.obtnsa , Chessm., 
the appendages are so well developed that they are as a rule equal in size to 
the whole lamina. 
Excellent figures of the leaves are given by Cavanilles (Ic., v, PI. 
ccccxiv), and by Hooker in Bot. Mag., t. 4056 , for C. sagittata ; by Hooker 
for C. novae-zelandiae (EL Nov. Zeal., t. 6 ), and by Gay for C. andicola 
(Atlas Bot., PI. ii). 
The figures interspersed in this paper also illustrate the point, and are 
included to show the method of development of the more remarkable and 
distinct appendages of some of the other species. But for the existence 
of these simple forms of inflexion the morphological significance of the 
appendages in C. appendiculata y Pers. (see Deless., Ic., t. 43 ), C. dioneaefolia , 
Hook. f. (see P'1. Antarct., ii, Pi. lxxxiv), C. alata , A. W. Hill (Figs. 3 and 
4 ), and C.phylloptera y A. W. Hill (Eig. 7 ), would be far from obvious. 
The next stage in the development of the appendage from the simple 
upturned lobe is shown by C. introloba , F. Muell. (Fig. 8 ), and C. involuta y 
A. W. Hill (Fig. 2 ). In the former species the axis of the fold is no 
longer at right angles to the petiole, but has become inclined obliquely 
to that organ at an angle of 45 0 . The appendages are also considerably 
elongated towards the apex of the leaf, being produced forward over the 
surface of the lamina, and, as the fold lies partly open and is not pressed 
1 Urban remarks (Eng. Eot. Jahrb. xxxvii, p. 401) that the genus Caltha is represented in 
collections from Colombia, Ecuador, and Chile by a widely distributed species, C. sagittata , Cav. 
( = andicola , Gay). He thus failed to appreciate the morphological peculiarities of the leaves of the 
Northern Andine forms, and is also incorrect in reducing C. andicola , Gay, from Chile to the Magel¬ 
lanic species C. sagittata , Cav. 
