ii6 FLORA AND SYLVA 
mata branched. Although discovered 
by Hohenacker in 1 8 3 8 it was not until 
the year 1902 that it was introduced to 
cultivation,as reported in the Gardeners' 
Chronicle oi 26th December 1903. It 
was collected in Russian Talysch, at the 
comparatively low elevation of 1,000 
feet, growing in the shade of low bushes. 
At first it promised to keep up its re- 
putation of having the longest flower- 
ing period of any known Crocus — from 
September to March ; but now that it 
is settling down it is becoming a late 
autumnal flower, and almost all of mine 
showed bloom this last season before 
December. It has proved hardy in shel- 
tered nooks in the rock-garden here, 
but I have kept the greater part of my 
stock in a cold frame, where the large 
flowers open freely and last longer than 
those of most Croci. 
The typical form is pure white with 
a rich orange throat, but the flowers 
vary much in the amount of orange 
appearing externally; in the var. lila- 
c/nus the segments are of a delicate rosy- 
lilac shade, unlike that of any other 
Crocus. For pot-culture in a cold frame 
or Alpine-house, it is a very valuable 
plant, flowering during late November 
and early December when flowers of 
that class are few. It appears to possess 
a vigorous constitution and should take j 
the place of C. Boryi^ which, owing to 
its delicacy , has almost disappeared from 
English gardens. In the engraving the j 
corm-tunic appears as though of fine | 
parallel fibres, but this is the result of! 
the method employed, the living tunic \ 
showing a perfectly smooth membrane | 
of reddish-brown colour. 
C. dahnaticus, — One of the earliest 
to bloom of the vernal species ; by mid- 
January I generally have its first flowers 
open. It is unfortunately a very rare 
plant in English gardens, though pos- 
C ROC us Dalmaticus {Visiani). 
Engraved for Flora" from a drawing by Mr E. A. Bowles. 
sessing a good constitution and seeding 
fairly well. The plant generally sold as 
dalmaticus is the Dalmatian form of C. 
biflorus^ rightly known as var. Weldeni 
— a mistake that can be easily detected 
at any time, for the true dalmaticus has 
