Junk 22. 
THE COTTAGE GARDENER. 
213 
every side. London and its southern environs lying 
maplike to the north and west; the tall masts of the 
shipping, and the smoke from the steamers on the 
bosom of the distant Thames, visible north-eastward on 
a clear day; while to the south, beyond Croydon and 
the Epsom downs, the far view is closed by the long, 
wavy range of the Surrey Hills. So material a change 
of site made necessary alterations no less important in 
the form of the Palace itself. According!}', the arched 
roof suggested by Sir Charles Barry was substituted for 
the flat one of the old building, increasing the height by 
44 feet, while the disproportionate length was reduced 
by 240 feet. Instead of a single transept there are 
now three—one at each end of the nave, of the same 
dimensions as that of the old building—and a central 
one of the majestic height of 104 feet and 120 feet span. 
The latter crosses the roof of the nave, rising high 
above it; the north and south ones meet it only with 
low square towers at the intersections. Owing to the 
slope of the ground, the Palace has an addition or base¬ 
ment story on the park side, forming a curious sort of 
tunnel in its interior (occupied by the warming appa- : 
ratus), and giving a most imposing elevation to the j 
facade. From the north and south ends advance, on 
the park side, glass wings, terminated by oblong towers; 
the southern one communicating with the new station ! 
of the Brighton Railway within the grounds. On this 
side also, which is in fact the front, the arched ends of 
the transepts are recessed to the depth of twenty-four 
feet; leaving an open arcade, covered only overhead, 
under each, which has a most picturesque effect, throw¬ 
ing the glazed ends into deep shadow. Viewed, therefore, 
from the park, the Crystal Palace forms three sides of a 
parallelogram, the longest of which only is visible from 
the road at the rear.” 
YOUNG PEAR TRAINING. 
(Continued from page 192.) 
In my last paper I concluded the Pyramid and Table- 
trellis forms. I now proceed with the next in order,— 
The Umbrella Mode. —I have coined this title to 
express that form which consists of one straight stem 
with the shoots trained downwards from a centre. This 
mode of training is tolerably well adapted to some kinds 
—principally those with slender and supple shoots; 
those of a gross and stiff habit, and inclined to per¬ 
pendicular growth, are rather inapt; they are constantly 
at war with this principle. However, we must try and 
deal with them. A tree of this kind, when selected, 
should have a clean, strong, self-supporting stem, of 
about three to five feet in height, according to the taste 
and designs of the cultivators, and should possess from 
four to six or more shoots at the summit; but these shoots 
should be of the last year's wood, or such as will readily 
bend. Of course, some kind of trellis will be requisite, 
at least for half-a-dozen years at first. I have known 
many fine trees trained on supple sticks out of a coppice 
—such as Oak or Hazel; but it is far better, and looks 
by far more artistic, to use an umbrella-form of wire, 
such as we have all seen applied to the training of Tree- 
Roses, but, of course, in shape and size adapted to the 
Pear. 
The training rods may radiate from the centre, and 
in outline may be the segment of a circle. It would, 
doubtless, however, be an improvement on the old plan 
to adopt two circles of wire at the top, from the outer 
of which the radiating lines might proceed. Thus, one 
circle might be established at four inches from the apex, 
and another (parallel) at about five inches from that; and 
from the latter may radiate the downward curving wires, j 
This will, in a measure, obviate the crowding which is I 
sure to take place near the top, owing to the acute angle 
from which the radiating wires proceed when they start 
immediately from the apex. The wires may start at 
about eight inches distauce from the outer circles, and 
by the time they reach the ground, or nearly so, they 
w r ill be about one foot apart. It is best to let them all 
terminate on a horizontal rod, at about four inches or 
so above the ground level; this gives strength and con¬ 
sistency to the whole, and keeps the shoots and Pears 
from being splashed. 
On such wires, then, or bended sticks in a similar form, 
the first leaders must be carefully trained as soon as 
they can be handled, getting one on each as soon as 
possible; but this is not always accomplished in the 
first year. If there be but a limited number of shoots— 
say, four—they should be divided equally around. 
The first training will, of course, take place during 
the winter pruning; and if the head is deficient of 
shoots, one, two, or more if necessary, of the very 
strongest shoots must be pruned back to about four or 
five inches, in order to multiply the leaders for the next 
year. If the wire circles are established, they may be 
pruned back to the outer circle. Throughout the 
summer the growing shoots must have attention, more 
especially during June and July; and in another year 
or two they will need disbudding, pinching, &c., as other 
trained Pears—of which I will speak when dealing with 
more mature trees. 
The Old Trellised Espalier. —This form is now 
well nigh out of date in the Pear way; at least, in our 
principal gardens ; but I really cannot say why, except 
that newer and more fanciful modes have thrust it aside, 
for there is much merit in the form, if well carried out, 
as to certain kinds. I suppose that the frightful amount 
of breast-wood, and the consequent barrenness of the 
old trees of former times, which proved so unprofitable 
in many old gardens, gave people au idea that the fault 
was in the system. Not so, however; bad management, 
and ill-selected kinds, have cast a prejudice on this mode 
of training, which fairly deserves a revival. I have an 
intention, some day, to convey my ideas, in this work, on 
what a kitchen-garden should be in first-rate style; and 
when I do, I will endeavour to work in this espalier 
mode in part. 
I may just remind our readers, that such trellises 
were, for the most part, composed of wood, and sometimes 
rough sticks, which were constantly falling into decay; 
that they were generally placed at the back of a marginal 
border, of some three or four feet, and that they wore 
from about five to six feet in height. They consisted of 
upright posts, at about six feet distances, and horizontal 
bars or rails, at about eight inches apart, in parallel 
lines. The whole of the branches were, of course, in¬ 
tended to be trained horizontally, but such a systematic 
course of procedure was but too often departed from as 
the trees advanced in age, and hence confusion and a 
disrelish of the practice. Now I do not desire, for a 
moment, to insinuate anything against the gardeners of 
that day; I am well aware that many clever and first- 
rate men had espalier lines of this character. The fact 
is, that they were ignorant of the causes which led to 
such confusion and loss; they did not, at that period, 
fully recognise the necessity of placing the roots of their 
fruit-trees under control, more especially of the Pear. A 
deep and rich border must be made ; everybody did it; 
and this long passed as a justification of a practice 
having a far greater tendency to produce wood than 
fruit. I well remember a joke that passed about, in my 
younger days, in the neighbourhood of London. A cer¬ 
tain gentleman, who was exceedingly desirous of going- 
a-liead, unloosed his purse-strings with much latitude, 
and employed a deep and rich border-man to establish 
him a fruit-garden. It was done; and two or three years 
passing away without produce of any character, the gen- 
