n 6 C. H. Ostenfeld. 
(F. Schultz), Marss. from the river at Monsal Dale, Derbyshire, and 
R. peltatus Schrank, f. truncata Koch from a small rivulet at the 
Upper Lake, Killarney. 
NyMPH/EACE/E. 
Nymplicea (vel Castnlia) alba L., var. occidcntalis Ostf., nov. 
var. 
Divert a specie : Nervis loborum folii primariis si in lineam 
rectam producas, convergentibus; receptaculo ad insertionem 
subangulato ; fructus parte summo cicatricibus staminum desti- 
tuta ; seminibus majoribus (3—3-5 mm.) 
Dijferta subspecie N. Candida Presl. : Granis pollinis verrucoso- 
papillosis; carpellorum numero majori; radiis disci stigmatici 
unicuspidatis flavis ; germine sub disco minus angustato ; fructu 
globoso vel depresso-globoso (5:7). 
Hab. Scotia (Perthshire: Loch Cally prope Dunkeld) et 
Hibernia (Galway: Craigga More). 
The small waterlily found during the excursion, first in Scot¬ 
land and later in Ireland, was provisionally identified by me as 
Nymplicea Candida Presl., and this determination has been followed 
by Druce in his paper on the “ Foristical Results” (l.c. p. 324, 
New Phytologist), where he quotes the description of N. Candida 
Presl. from H. S. Coward, The Waterlilies, (Carnegie Publ., 
1905). This description, compiled from different sources, and also 
the better descriptions in Caspary’s different papers, do not wholly 
Nearly ripe fruits in longitudinal section. Two-thirds nat. size. 
A. Nymplicea alba (from Denmark), alcoholic material. B. N. alba var. 
occidentalis (from Craigga More, Co. Galway), soaked herbarium material. C. 
N. Candida (cultivated in the Botanical Garden, Copenhagen), alcohol 
material. 
agree with our Scottish and Irish specimens. I was, therefore, 
in some doubt as to the correctness of my identification, and have 
now examined my specimens more closely and have compared them 
with numerous specimens of N. alba and N. Candida from Sweden 
and Eastern Prussia, Bohemia, etc. The result is that the British 
plant is not the true N. Candida , to which it has a strong external 
resemblance. The above given diagnosis will show that it is inter¬ 
mediate between the two forms; having some characters of N. 
alba, others of N. Candida. Ascherson and Graebner (FI. Nor- 
dostdeutsch. Flachland, 1898-99, pp. 319, 320) also mention, that 
while the two forms are well distinguished in Prussia, they have often 
met with forms in Pomerania and Brandenburg : “ die sich mehr 
u. mehr der N. Candida nahern, ohne ihr jedoch gleich zu kom- 
men, so dass hier eine scharfe Grenze nicht vorhanden ist. 
Wir haben es hier augenscheinlich mit einer in der Ausbildung 
begriffenen Art zu thun.” I think something like this is the case 
