DIMORPHISM IN ERANTHEMUM. 165 
the po greyish-white, about 2 lines long; tubular, with a 5-toothed 
limb. e minute corollas in this species usually wither on the deve- 
loping ai so that it is scarcely possible for them to be crossed with 
pollen from other flowers. 
t may not be out of place to add a note on Leersia oryzoides. I 
have long entertained a hope that cultivation under such extremes of 
temperature as nage naturally afforded in the plains of India might induce 
the development of open and normal flowers on this inveterate pro- 
ducer of Lor and selí- it flowers. Mr. Darwin was kind enough 
to send me seeds for ex ent. "These ed me in Octo 5, 
and were at once sown ey germinated freely, and yie in beta 
progeny which began to flower in Mare 6, continuing 
ess or more to the close of the rainy season. In September they yielded 
an abundant crop of seeds, all of which, however, were the produce of the 
characteristic closed flowers. Year after year have I watched it in the 
r. Darwin. Thus s cue confirmed is the habit of Bea those 
minute abnormal flowers! Forlorn as it may appear, however, I still 
think that by successive changes in the modes of culture, the normal 
flowering habit of this interesting little genus will be restored. 
entertain this view under the conviction that the tendency to produce 
wth, and thus more or less liable to ‘submergence a e t 
sterility of all open flowers in rainy weather, the production of 
closed flowers equally fertile—as I kn they are—in w and 
a great gain. Again, in the several 4canthacee oe 
in alr, Is a very great gal n 
enumerated, we have very clearly seen that the production of the 
forms of flowers is largely if not wholly dependent on pni. psi 
conditions. In other plants nature effects the same end—multiplication 
f the species—by other modes, e.g. the pa pet "a of mountain 
icis which are normally reproductive in the plains, the germination 
Mangroves and other such plants while attached to the parent 
within the tidal range, and their falling off un rminated (as is very 
generally the case) when growing beyon d “its range on merely moist lands. 
Reflect also on Agaves, which are - dispersed over India, and found in the 
most dissimilar habitats. All of these i in hot and arid districts produce 
and denim we have well-mar rked viviparous vali 
sort of qua nia, ve e mode of eiue (as in the leaf-bud ed 
hnic of apparently arrested- ovules; ¢.g., Agave and Kalanchoe) to a 
