I ( 
the rummer months. Profeffor Beckmann, in the third 
volume of his Hiftory of Inventions^ has proved clearly 
that the ancients were well acquainted with what ferved 
the purpofe of ice-houfes. ‘‘The art (fays he) of pre- 
ferving fnow for cooling liquors during the fummer, in. 
warm countries, was known in the earlieft ages. This 
practice is mentioned by Solomon, Proverbs xxv. 15. and 
proofs'of it are fo numerous'in the works of the Greeks 
and the Romans, that it is unneceffary for me to quote 
them, efpecially as they have been collected by others. 
How the repofitories for keeping it were conftructed, we 
are not exprefsly told; but it is probable that the (now 
■was preferved in pits or trenches. When Alexander the 
Great befieged the city of Petra, he caufed thirty trenches 
to be dug, and filled with fnow, which was covered with 
oak branches ; and which kept in .that manner for a long 
time. Plutarch fays, that a covering of chaff and coarfe 
cloth is fufficient; a'nd-at prefent a like method is pur¬ 
ified in Portugal. Where the fnow has been collected in 
a deep gulf, lome grafs or green fods, covered with dung 
from the Iheep-pens, is thrown over it; and under thefe 
it is fo well preferved, that the whole fummer through it 
is fent the diftance of iixty Spanifh miles to Lilbon. 
“ When the ancients, therefore, wifhed to have cooling 
liquors, they either drank the melted fnow, or put fome 
of it in their wine, or they placed jars filled with wine in 
the fnow, and ffiffered it to cool there as long as they 
thought proper. That ice was alfo preferved for the like 
purpofe, is probable from the teftimony of various au¬ 
thors ; but it appears not to have been ufed fo much in 
warm countries as in the northern. Even at prefent fnow 
is employed in Italy, Spain, and Portugal; but in Perfia 
ice. I have never any where found an account of Gre¬ 
cian or Roman ice-houfes. By the writers on agriculture 
they are not mentioned.” 
In the choice of a fituation for an ice-houfe, the prin¬ 
cipal regard Ihould be, that of a dry fpot of ground ; for, 
wherever there is moifture, the ice will melt; therefore in 
all Itrong lands, which detain the wet, there cannot be 
too much care taken to make drains all round the build¬ 
ing to carry oft’ all moifture ; for, when this is lodged 
near the building, it will occafion a damp there, which 
will always be prejudicial to the keeping of the ice. The 
next confideration mult be, to have the place fo elevated, 
that there may be defcent enough to carry off whatever 
wet may happen near the building, or from, fuch portions 
of the ice as may occalionally melt; alfo; that the place 
be as much expofed to the lfin and air as poflible ; and 
not placed under the drip or in the fhade of trees, as hath 
been too often praftifed, under a falfe notion, that, if it 
Ihould be expofed to the fun, the ice will melt away in 
fummer; for this never can be the cafe where there is 
fufficient care taken to exclude the outward air (which 
mult always be regarded in the building of thefe houfes), 
as the heat of the fun can never penetrate through the 
double arches of the building, fo as to add any warmth 
to the included air; but, when the building is entirely 
open to the fun and wind, all damps and vapours will 
thereby be removed from about the building, which can 
never be kept too dry, or free from moift vapours. As 
to the figure of the building, that may be according to 
the fancy of the owner; but, for the well into which the 
ice is to be put, a circular figure is the molt convenient; 
the depth of the well, as alio the diameter of it, mult be 
proportioned to the quantity of ice wanted ; but it is al¬ 
ways beft to have enough ; for, when the houfe is well 
built, it will keep the ice for two or three years; and 
there will be this advantage in having it large enough 
to contain ice for two.years confumption, that if a 
mild winter fhould happen, when there is not ice to be 
had, there will be a flock to fupply the want. 
If the quantity wanted is not great, a well of fix feet 
diameter, and eight feet deep, will be large enough ; but, 
for a large confumption, it fhould not be lefs than nine 
ten feet diameter, and as many deep; where the fitu- 
! E. 723 
ation is either dry chalk, gravel, or fand, the pit may be 
entirely below the furface of the ground; but, in ftrong 
loam, clay, or moift ground, it will be the beft way to 
raife it lb high above the furface, as that there may be no 
danger from the wet. At the bottom of the weil there 
fhould be a fpace left about two feet deep, to receive any 
moifture which may drain from the ice, and a fmall un¬ 
derground drain fhould be laid from this, to carry off the 
wet; over this fpace of two feet, fhould be placed a ftrong 
grate of wood to let the moifture fall down, which may 
at any time happen from melting of the ice. The fides 
of this well muft be walled up with brick or ftone, atleaft 
two feet thick ; but, if it is yet thicker, it -will be better, 
becaufe, the thicker the walls are made, the lefs danger 
there will be of the well being affe< 5 ted by any external 
cauf'e. When the well is brought within three feet of the 
furface, there muft be another outer arch or wall begun, 
which muft be carried up to the height of the top of the 
intended arch of the well; and, if there is a lecond arch 
turned over from this well, it will add to the goodnefs of 
the houfe; but, if not, then the plate into which the roof 
is to be framed muft be laid on this outer wall, which 
fhould be carried high enough above the inner arch, to 
admit of a door-way in, to get out the ice. If the build¬ 
ing is to be covered with flates or tiles, there fhould be a 
thicknefs of reeds laid under, to keep out the fun and 
external arr; if thefe reeds are laid two feet thick, and 
plaftered over with lime and hair, there will be no danger 
of the heat getting through it. The external wall need 
not be built circular, but of any other figure, either 
fquare, hexangular, or oftangular; and, where it Hands 
much in fight, may be fo contrived as to make it a good 
objeft. 
The aperture of the mouth of the well need not be 
more than two feet and a half diameter, which will be 
large enough to put down the ice, and there fhould be a 
ftone fitted to llop this aperture, which muft be clofed up 
as fecure as poflible, after the ice is put in, and all the 
vacant fpace above and between this and the outer door 
muft be filled clofe with barley ftraw, to exclude the air ; 
fo the door to enter for taking out the ice fhould be on 
the oppofite fide ; and this door fhould be no larger than 
is abfolutely neceffary for the coming at the ice, and muft 
be ftrong and clofe to exclude the air ; and at five or fix 
feet diftance from this another door fhould be contrived, 
which fhould be clofely fhut before the inner door is 
opened, whenever the ice is taken out. 
The building, being finifhed, fhould have time to dry 
before the ice is put into it; for, when the walls are green, 
the damp of them frequently melts the ice. At the bot¬ 
tom of the well, upon the wooden grate, fhould be laid 
fome fmall faggots ; and, if upon thefe a layer of reeds is 
placed fmooth for the ice to lie upon, it will be better 
than ftraw, which is commonly ufed. In the choice of 
the ice, the thinner it is, the better it may be broken to 
powder; for, the fmaller it is broken, the better it will- 
unite when put into the well. In putting it in, care muft 
be taken to ram it clofe; as alfo, by laying ftraw, to al¬ 
low a vacancy all round next the wall, of about two 
inches; this is to g-ive paffage to any moifture which 
may be occafioned by the melting of fome of the ice on 
the top, which, if pent up, will melt the ice downward.. 
When the ice is put into the well, if there is a little falt- 
petre mixed at every ten inches or a foot thicknefs, it will, 
caufe the ice to join more clofely into a fblid mafisfi The 
ice being all incorporated into one folid body, there muft 
be a crow or other iron inftrument always in readinefs to 
break it up; taking out no more at a time than is wanted. 
Mr. Miller propofes to make ufe of the ice-houfe as an 
ornamental building; that however is feldoni done; and 
it is generally placed in a fequeitered fpot, on the fide of 
a hill or Hoping ground, the bale of which is lower than, 
the bottom of the weil; the outfide being well banked 
up with earth, the better to keep out all air and heat, and 
the embankment being neatly covered- with turf, 
leg- 
