140 Select Carnations, Picotees, and Pinks. 
flowering stems, had a tendency to develop a “tree ” 
habit, throwing out shoots from stems that con- 
tinued to elongate, and which could not be layered 
in the ordinary way, but had to be perpetuated in 
the same manner as a tree Carnation. This was 
discovered when propagating seedlings, which 
flowered on all the basal shoots and had to be per- 
petuated by means of pipings from the flower stem. 
The above was confirmed by the experience of 
one of his correspondents. It was also the ex- 
perience of the latter that layers constantly gave 
rise to plants with an abundance of grass at the 
base of the stems. Another striking fact was that 
pipings taken from the flower stems bloomed ten to 
twenty days later than plants from layers; and the 
period of flowering was retarded till September or 
October, when pipings were taken late in the sea- 
son. This was more decidedly emphasised when 
the plants furnishing the pipings were of weak 
soa It also occurred to this correspondent that 
ne exhibition varieties might be systematically 
propagated in this way with the view of getting 
plants to flower in mid-winter or early spring. 
Mr. C. W. Ward, the author of “The American 
Carnation” also had a curious and interesting ex- 
perience. He was in the habit of raising large 
numbers of seedlings annually and growing them 
like Tuberous Begonias in the open field for the 
purpose of selecting the best for further trial. One 
autumn, after all the promising varieties had been 
placed under glass, he discovered a plant with a 
large brilliant scarlet flower at the end of a long 
trailing stem. It was lifted from the frozen ground 
and propagated by bud sports till it assumed a 
dwarf habit and became one of his most profitable 
varieties, both commercially and as a parent from 
which many of his best productions have since 
descended. 
He further observed that like produced like when 
selecting cuttings for the propagation of named 
sorts. Strong cuttings from vigorous _ plants 
repeated the characters of their parents, and those 
from diseased, stunted, or weak plants repeated 
themselves. Buds giving rise to colour variations 
could be fixed by continued selection of cuttings 
from the sport. By repeatedly selecting cuttings 
from abnormally leafy shoots, showing little inclina- 
tion to bloom, plants were obtained that failed to 
bloom after being cultivated for two years. Several 
