COAL-TAR AXD WATER-GAS TAR CREOSOTES. 29 
The flask in position for distillation is supported on an asbestos 
board in which a hole has been cut almost as large as the great 
diameter of the bulb. The outline of this opening is irregular to per- 
mit the flame to play about the bulb. No wire gauze is used, for 
there is no danger of breakage so long as the flame does not come in 
contact with dry glass. The portion of the bulb above the asbestos 
board and below the Hempel column is protected from drafts by an 
asbestos box. No protection is given to the Hempel column unless 
the condensation becomes so great that the column begins to fill, and 
then it is surrounded by an asbestos box sufficiently large to prevent 
any portion of it coming in contact with the glass. The condensers 
used are ordinary glass tubing about one-half inch in diameter, 
drawn down at one end to about one-fourth inch and flanged at the 
other to receive a cork stopper. The length is approximately 8 
inches. This has been found to be sufficient to condense all of the 
lighter oils but not sufficiently long to cause much solidification of 
the higher distillates in the tube. If solidification should occur, it is 
always melted out before the fraction is taken. 
The thermometers used in this work were made of Jena borosilicon 
glass and were filled above the mercurv column with carbon dioxide 
at a pressure of approximately 15 atmospheres. They read from 180° 
to 550° C. and were standardized either by the Bureau of Standards 
or by the German Physikalisch Technische Reichsanstalt. Correc- 
tions were made for the slight inaccuracy of the thermometers and 
also for the emergent stems. This latter correction is not usually 
made in creosote distillation, and would not be necessary if a standard 
thermometer were used for all such distillations and if the length of 
the emergent stem were the same in all cases. At the time this work 
was started no standard theremomter for creosote distillation had 
been adopted or even proposed by the various associations interested 
in wood preservation. The correction was necessary, therefore, as 
only by this method could the data here presented be compared with 
data in the collection of which thermometers of different lengths had 
been used. 
For those who are not familiar with the emergentstem correction, 
the following explanation of its meaning and use is given. Mercurial 
thermometers are usually standardized with the whole of their 
mercury column at the same temperature as that of the bulb. This 
is possible only when the total length of the thermometer is bathed 
in the heated vapors or liquid. If this is impossible, then the thread 
of mercury is cooler than the bulb, and an error is introduced which 
makes the observed reading too low. If the difference between the 
temperature of the stem and bulb is small, this error is negligible; 
but if the difference is large, and the length of the emergent stem is 
great, the error becomes a factor of considerable importance, and 
