204 
advance in growth they should be pegged down to the surface 
at some of the joints, and covered here and there with a little 
soil, to cause them to make young roots; and when once the 
plants become thoroughly established they will grow rapidly 
into a dense mass of interwoven creeping stems. Amongst 
these some light peat may, occasionally, be mixed; and if it 
be desired that a considerable space should be covered with 
the plants, they may be divided in April, into small patches, 
and taken up with the soil about their roots and then planted 
two feet apart; and by subsequently attending to the pegging 
down of the shoots, they will cover, in little more than a year, 
all the spaces left between the plants. There are what are 
considered to be two varieties in cultivation ; that from North 
America is said to be of freest growth, but a shy flowerer; 
whilst the other — a Scotch variety, is said to flower freely but 
is of slower growth. After cultivating both sorts, for several 
years, I find them equally luxuriant, and to flower equally well, 
when treated as I have directed. Generally, in the month of 
June they flower abundantly, and prove quite hardy. 
196 Layers, Rooting. To the late Thomas Andrew Knight, 
the vegetable physiologist and horticulturist are more indebted 
than to any individual of the present century. His numerous 
experiments have determined questions, and exposed facts pre- 
viously un-thought-of, we shall, therefore, occasionally select 
from his papers which have lately been published by his family, 
with a sketch of his life, forming a single, but highly interes- 
ting, volume. The following experiment on Layers, will 
suggest to the propagator of shrubs new modes of operation. 
When a layer is prepared, and deposited in the ground, the 
progress of the sap, in its descent towards the original roots, is 
intercepted upon the side where the partially detached part, or 
tongue, of the layer is divided from the branch ; and this in- 
tercepted sap is, in consequence, generally soon employed in 
the formation of new roots. But tliere are many species of 
trees which do not readily emit roots by this mode of treat- 
ment ; and I suspected that, wherever roots are not emitted by 
layers, the sap, which descends from the leaves, must escape 
almost wholly through the remaining portion of bark, which 
connects the layer with the parent plant. I therefore attempt- 
ed, in the last and the preceding spring, to accelerate the 
