32 NARCISSUS POETICUS AND ITS ALLIES 
which resemble N. poeticus pacreaeys in stamens and perianth, 
ae a pe corona; N. exertus and N. poetarwm, while agree- 
ing with N. radiiflorus in their subequal stamens, have the flat 
ess - N. oeticus; and the differences in the fruits, mags 
are not correlated with those of any of the other organs. But 
there seems no more practical arrangement of these Slane than 
the selection of Koch’s two species, which are really widely 
different, as types of two series, Eu-poeti ci and Radziflori, of 
which the unequal stamens, with imbricated perianth-segme ents, 
of N. poeticus, and the subequal stamens, with cuneate-basei 
perianth-segments, of N. radiiflorus are regarded as the essential 
features. a series is then readily subdivisible by its flat or 
cupular coron 
There is some evidence that this arrangement is not only 
practical but natural in the geographical distribution of these 
plants, for the rer series, Hu-poetici, extends across Southern 
a ad from Spain through Southern France to Italy and Greece, 
while pt second, Radziflori, with the possible exception of 
N. poetarum, _— native country is unknown, favours more 
northern and eastern regions, ranging from Switzerland across 
Austria to Tra sate eat and the Balkans. In the Swiss Alps, 
where the habitats of the two series meet or bee we got the 
anomalous species N. exertus and N. verbanensis on the rth 
the Greek N. hellenicus is a rare outlying species, showing 
marked peculiarities as might be expected. Among the Radiiflort, 
the arian a N. stellaris is probably generally distributed 
— the Tyrol to Transylvania, while the earlier N. radiiflorus 
s to occur in the neighbourhood of the capa sree 
Stak eastwards from Trieste to Servia and Monten 
different forms, and to M. Hisieenk of the Boissier Herbar as 
Gen for living specimens and detailed information 
reapecting ei Alpine Narcissi, I am also indebted for valuable 
suggestions to Dr. O. Stapf. 
As the work of early writers is of — interest in this 
genus a come pree-Linnean synonymy will be given. Exsiccata 
will not be uoted owing to the general obbioealics of floral 
characters and the doubt attaching to many sets that have been 
sent out. It may be suggested here that, in collecting these 
Narcissi, notes should be made of the form and colouring of the 
— and the arrangement of the pennene. mee ig of the 
ould also be whenever pos ollowing 
aire: have been drawn up from ee. ce axocpt where 
