iculty, and have been obliged to modify my plan
in various respects:

In the first place, wooden tanks, with the top covered with sand, will
not give off heat sufficiently to keep up growth in houses of this size
during extremely cold weather. By protecting the houses with shutters,
this difficulty may be obviated. Crowding the fire, and raising the
water in the tanks to a high temperature, is a more objectionable
remedy. In this way the bottom heat is too strong. But my most serious
difficulty has arisen from excessive humidity. I put three inches of
sand over the whole slate surface of the tanks, using a part for
cuttings, and the rest, (say 100 running feet of the three feet wide
table), for standing pot plants upon the surface of the sand. The plants
dried rapidly, and required watering every morning. The result was, that
in watering the plants, and of course the sand on which they stood, to
some extent, it was like pouring water upon a flue, or upon hot pipes: a
constant steam was given off; all the moisture in the sand was rapidly
converted into steam; so, also the water in the pots was quickly
expelled. In order to heat the house sufficiently, the bottom heat
became too strong, and the plants were in too direct contact with it. In
cold days the house was in a perfect fog. It was ruinous to the plants.
The remedy was simple: more heat must be allowed to escape from the tank
into the house, without coming in contact with the sand-bed, and the
moist earth of the plants. Another slate floor was laid, an inch above
the tank slate, on which to put the sand and stand the plants. This hot
air chamber opens into the house on the back and front side of the tank.
Thus the whole radiating surface of the top of the tank may be directed
into the house, or may be confined as bottom heat, as may be found
necessary. By this plan, excessive humidity may be entirely obviated,
and the heat completely controlled, as wanted."

HOT WATER PIPES.--It is generally conceded, among practical men, that
the circulation of hot water in iron pipes is the best known method of
heating plant houses. The property which heated water possesses of
retaining for a considerable length of time its heat and transmitting it
to the pipes at long distances from the boiler, renders it a most
effective agency for such purposes: A perfect control of the moisture of
the atmosphere, by means of evaporating pans attached to the pipes;
entire freedom from deleteri