128 ON A PROCESS OF FRACTIONAL CONDENSATION. 
except that they required too frequent renewal. I have found the cloth cove 
with vulcanized caoutchouc preferable to the common caoutchouc tubing. In 
smaller sizes of apparatus I have the end of the worm itself project far enough f 
the bath to connect directly with the retort by means of a perforated cork, with 
the use of an additional connecting tube. 
The upper end, 4, of the elevated worm is brought out through the side of the t 
at a point about three inches below the top; so that, when working with a low t 
perature of the bath, the worm may still be completely covered with oil, and also 
sufficient space above the worm for the expansion of the oil when higher temp 
tures are employed. To avoid contaminating the atmosphere of the laboratory v 
the disagreeable fumes which are given off, in large quantity, from such a mast 
heated oil, the top of the bath is tightly closed with a sheet-iron cover, from whic. 
small funnel, a, Fig. 1, conducts these fumes to a chimney. 
In the larger apparatus, the vapors which succeed in passing through the hea 
worm are conducted downward into a cooled worm contained in a bath of wate 
Fig. 2, and the liquid product is collected in the receiver, 4. The cold bath, #, conte 
two condensing worms, — one for each apparatus, — and is large enough to conde 
for both without the necessity of renewing the water. I have represented two ap 
ratuses combined, as it will be found more economical of time to operate with twc 
once. In the smaller apparatus, for the table, a Liebig condenser may be convenien 
substituted for the cold worm, as shown in Fig. 1. 
For collecting liquids which boil below the common temperature, when such 
present, I attach a refrigerator, B, Fig. 2, which is provided with two block-tin ¢ 
densing-tubes, — one for each apparatus. These are bent in a zigzag form, < 
attached to the inner sides of the refrigerator. The lower ends of the tubes ext« 
through the end of the refrigerator far enough to form a convenient connection w 
the second receiver, /, Fig. 2, which communicates with the first receiver, 4, by me 
of the glass tube, m. 
In order to successfully collect and condense the vapors of such extremely vola 
liquids as are now under consideration, it is of course indispensable that the appara 
should be constructed with very tight joints; and for greater convenience, but m 
especially to prevent breakage, such of the joints as require to be frequently tal 
apart should be made flexible. A very convenient and perfectly tight joint of { 
kind may be made as follows:— the short stationary tube, x, in the cork of - 
receiver, #, Fig. 2, is made with the opening somewhat divergent upward ; the enc 
of the worm is enough smaller than the inside diameter of the upper end of the tu 
