ON A PROCESS OF FRACTIONAL CONDENSATION. 127 
time, and fuel in heating the bath, as it diminishes the quantity of oil required to 
cover the worm. It is made to extend to within about three inches of the bottom of 
the bath, and large enough to fill the greater part of the space in the centre of 
the coil. The bath and interior vessel are both made of sheet-copper, with joints 
brazed so that they will bear a high temperature. I generally use, also, copper 
worms, especially in the earlier distillations, the quantities then operated upon being 
larger, as such worms are conveniently procured, and not liable to break. In the 
larger-sized apparatus, the tube of which the worm is made measures ten feet in 
length and half an inch in diameter. I have tried several lengths of worm and several 
diameters of tube, but not, as yet, with any special view of determining the precise 
proportions, in relation to the size of the retort, which would be best adapted to the 
purpose. There appears, however, to be nothing gained by increasing the length of 
the worm beyond what is required to reduce the temperature of the vapors to that 
of the bath. I have in use three sizes of apparatus: the largest has a copper worm 
10 feet long and } inch bore; the medium size, a worm 5 feet long and 2 inch bore; 
and the smallest size, for very small quantities, a worm 1 foot 6 inches long and 
3 inch bore. Each of these has been found to answer a good purpose. The distilla- 
tion may be conducted in a glass flask, or more conveniently in a glass retort of the 
form shown at d, Fig. 1 and 2. The body of this retort, as appears in the figure, is of 
the form of the corresponding part of the common retort; but which, in place of a 
long neck, has only a short tubulure, e, in the side, for escape of the vapors, and 
another tubulure, /, in the top, which contains the thermometer, and through which 
the retort is charged. 
In the larger apparatus the retort is connected with the lower end of the elevated 
worm by means of a glass tube of about the same diameter as the end of the worm. 
One end of this tube enters the retort at the lateral tubulure through a perforated 
cork, and the other end is joined to the end of the worm either by being firmly bound 
with a strip of cloth thickly covered with vulcanized caoutchouc, — such as is found in 
commerce, — or by means of a perforated cork, which is made to fit the ends of both 
tubes as snugly as possible, and then tightly pressed together upon the joint by means 
of an iron clamp, as shown at g, Fig. 2. This clamp is figured on a larger scale at x. As 
it-is highly important that all joints in the apparatus should be perfectly tight, inas- 
much as the least leakage, when continued a long time, would cause, in the aggregate, 
a serious loss of material, I would call special attention to the clamp joint as the best 
which I have tried. Before falling upon this device I had used exclusively the vulcan- 
ized caoutchouc joints, which were found to answer a good purpose, in most cases, 
