ion of cooking have been known as "thread," "soft ball," "hard
ball," "crack" and "hard crack." While the candy-maker has been
struggling with these unsatisfactory tests, the candy has kept on
cooking, perhaps until it has reached the next stage of temperature,
changed texture, and so become unfit for the use for which it was
intended.

The professional confectioner has long been able, by the use of a
thermometer, to determine just how hot his candy was and to remove it
from the heat at exactly the right moment. His thermometer, however, was
not only too expensive for the amateur, but also too long to be used
except in a large vat. There are now on the market, however,
thermometers that the amateur confectioner or small manufacturer can use
to advantage. Even when hardware dealers do not keep the thermometers in
stock, they can--and will--order them from their jobbers. The
instruments, of which there are several makes, are about nine inches
long, and sell for from one dollar to two dollars and fifty cents. Most
of these instruments register from about eighty degrees to three hundred
and eighty degrees Fahrenheit, although the range which interests
candy-makers most is only from two hundred degrees to three hundred and
twenty degrees. There are a few facts which the girl confectioner who
uses a thermometer should keep in mind.

Buy a thermometer which is guaranteed by its maker. See that the
markings are so well defined that they can be read easily. Before use,
the thermometer must be regulated to conform to the local altitude.
Place the thermometer in a kettle of water, heat, and let it boil for
ten minutes. If the mercury marks two hundred and twelve degrees, the
thermometer is correct as it is, but if there is a variation of two
degrees or more, allowance must be made. If water boils at two hundred
and ten degrees, two degrees must be subtracted every time the
thermometer is read. Then "soft ball" means two hundred and thirty-six
degrees, not two hundred and thirty-eight degrees, as it would had the
water boiled at the normal two hundred and twelve degrees, or two
hundred and forty degrees, as it would had the water boiled at two
hundred and fourteen degrees. Do not think that a variation of even two
degrees is too slight to count; absolute accuracy is essential.

For safety in transit, manufacturers often pinch together the case of
the instrument so that the scale-piece can not work loose and become
broken. A very