April 29, 1886. ] 
JOURNAL OF HORTICULTURE AND COTTAGE GARDENER. 
331 
of the first leaf or leaves above the cotyledons. In these seedlings it is 
curious to see how, whilst the primary roots descend vertically, the second¬ 
ary ones pass off horizontally. 
It is quite clear that the seed in Centunculus was not buried deeply, for 
thecaulicle is a minus quantity. It is clear also the soil for such a plant 
should be light, open, rich, well drained. Contrast this with the germina¬ 
tion of Primula reticulata, in which not only is the tigellum very long, but 
the two cotyledons are also raised on long, erect, or ascending stalks” as if 
the plant grew in the clefts of the rocks, and had a long distance to thrust its 
seed-leaves into the light and^air (fig. 57). The requirements of seedlings 
are, it need hardly be said, of the same character. W e all know the care 
that is requisite to secure the germination and the rearing of these delicate 
organisms. 
“ In the morning sow thy seed, and in the evening withhold not thine hand.’’ 
—Eccles. xi., 6. 
The mode of growth of the Androsaces is similar. They are, it is true, for 
the most part not annuals, but in their root-growth some of them (the 
casspitose species) follow the mode of annuals. The seedling plants have 
roots of the annual character, with a long radicle giving off numerous 
branches. The adult plant has little provision for storage, but consists 
of a dense tuft of leaves, from the axils of some of which proceed long 
slender runners, like those of a Strawberry, and which bear at their ends a 
tuft of leaves like the parent from which they sprung. From the under sur¬ 
face of this tuft proceed roots like those of the Centunculus, and, like them 
destined, not for any lengthened use, but only for a temporary purpose ; no 
long time indeed elapses ere the tuft throws out new runners and thus re¬ 
peats in another generation the process of its own genesis. It would seem 
from this peculiar mode of growth that the Androsaces speedily exhaust the 
area in which their roots spread, and having done so haste— 
“ To-morrow to frt sh woods and pastures new,’’ 
—Milton, Lycidas, 1.193. 
Whether this is not an indication of value to the cultivator I leave to 
others who have had more experience to decide. My own want of success 
with these plants cannot in fairness be attributed only to neglect of the 
indication ! 
PERENNIALS. 
In these, as for annuals, there is need for daily supplies during the 
growing period, and further, there is the necessity for replenishing the 
stores. It is necessary, therefore, to consider the root-growth of these 
plants from two points of view—that of food-collecting and that of food- 
storage. 
The actual absorption of water is, of course, effected in the same way in 
the roots of perennials as in those of annuals, but the perennial habit allows 
more time wherein to work, and frequently secures a wider root-range than 
is possible in an annual. In a perennial, cceteris paribus, the roots can 
travel further, or penetrate deeper in search of food, than in the case of an 
annual. 
In an annual the roots, as gardeners say, “ keep at home,” and there is not 
much necessity for a system of conduits to convey the water from its source 
to the stem ; but in p-rennials it often happens that the best food supply is 
at tome considerable distance from the stem, and the consequence is that 
the fibrous roots collect tbe liquid from the feeding-ground, and convey it 
in so many conduits to the stem. The thicker root-fibres have, as everyone 
knows, little or no power of absorption, that faculty being mostly limited to 
the thinnest extremities of the root-fibres, and to the root-hairs (where 
present). Of course the number, length, and degre ; of branching of the 
roots depend very greatly on the physical nature of the soil in which the 
plant happens to be growing— 
“Pinguibns hae terris habiles, levicribus illre.” 
But when due allowance is made for these circumstances each plant has 
more or less its own distinct character. The roots of most species of 
Primula, for instance, are very different from those of Androsace, and 
Fig. 61.—Alpine Auricula. 
indicate different requirements. But even in the same genus we get varia¬ 
tions in this respect. In the common Primrose and Polyanthus, in P. cash- 
miriana, P. capitata, P. amcena, P. Auricula, P. denticulata, P. nivalis, 
P. longiflora, P. cortusoides, ifcc., the roots are generally rather thick and 
fleshy, descending more or less vertically for some little distance without 
branching, and then giving off short nearly horizontal branches with few 
root-hairs, except in P. Auricula, in which, in some cases at least, the roots 
are covered with a velvety coat of hairs. Such roots are not surface-feeders, 
but are capable of penetrating to a considerable depth in search of food, 
while their Bucculent habit and reserve store of water obviate the necessity 
for that dense network of fibrous roots that other species present. Primula 
rosea (fig. 58), P. Kaufmanniana, P. involucrata, and P. cortusa Hatthioli, 
afford instances of this densely matted and comparatively superficial root 
development. They have few if any large root or conduit fibres; but, on 
the contrary, an intricate mass of fine fibrous roots, penetrating in every 
direction, and availing themselves, as it were, of every scrap of soil within 
reach. Surely we have here an indication of the necessity in cultivation 
of supplying these plants with ample depth and breadth of light, rich, 
moist, not to say wet soil. The utility of a mulch in hot dry weather is 
also indicated as a means of preventing the drying up of the surface roots. 
In some of the Androsaces we find provision for keeping the roots moist 
and free from the risk of drought. In seedling plants of A. elongata, for 
instance, the caulicle is very long, and the slender radicle descends vertically 
for a long distance without branching, and then gives off near its tip a leash 
of much-branched fine filaments. 
In Soldanella we have a tuft of rather thick fibres, which descend verti¬ 
cally into the soil, and are unbranched till near the points, where they give 
off numerous relatively short horizontal fibres. 
STORAGE AND BEST. 
In perennials we have specially to consider the arrangements for storage 
of food in convenient places for use when required, and the modifications of 
structure associated with the periodic alternations of active growth and 
relative rest. Of a Primrose, whether “ by a river’s brim,” or elsewhere, 
whether a yellow Primrose or one of other hue, it may be said that at one 
season its constant care is to increase its store, while at another its aim is to 
make all snug for winter. Of course the Primrose is not peculiar in these 
