888 
as to the value of the coal. The retort must be first got up to a bright red heat 
before the charge is introduced’ and must be maintained at this temperature 
during the whole of the distillation. 
The instructions are to take an average sample, buf there is very 
serious doubt in my mind if this is consistent with the limitations of 
“lumps the size of a walnut” and of approximately a kilo. In consider- 
ing this question’ 1 can see no reason why work on a smaller scale, 
selecting the sample more carefully, crushing the coal to smaller particles, 
and quartering to a sample of 40 or 50 grams should not give as good or 
even much more accurate results than would such approximate work done 
on the large laboratory scale recommended by Phillips. The work of the 
committee on uniform chemical analyses *® indicates that a gram, if 
properly selected, is an excellent sample of a ton of mineral, and a 
mineral is much harder to sample than coal. There seems to be no 
reason then why a carefully prepared sample of 40 or 50 grams is not 
ample, and this brings the size of the apparatus within the range of the 
ordinary chemical laboratory. 
The method, as given in the above quotation, is representative of many, 
and it seems that one of two things must happen in charging the retort, 
either some of the volatile hydrocarbons are lost before the apparatus is - 
closed, or the retort cools down in the charging. As a matter of fact 
these methods have been criticised by gas manufacturers as being apt to 
give low results both as to quality and quantity of gas. With a cool 
retort, which can be heated only gradually, the products of the aromatic 
series, such as toluene, naphthalene and anthracene, which ought to go 
into the gas, go into the tar. It is thought that both of these difficulties 
are eliminated and the yield more nearly like that obtained in a com- 
mercial way, by using the apparatus described below. 
Description of the apparatus.—Plate I, X and X’ (the latter is not shown in the 
figure) are two combustion furnaces, fitted together so as to form one continuous 
piece of apparatus. B is a piece of galvanized gas pipe of 1.5 centimeters inter- 
nal diameter capped at one end, and about 3 centimeters from the other end is a 
water jacket. Into the open end of the iron tube is fitted a rubber stopper, 
carrying a glass tube which delivers into the top of a Fresenius tower (C) filled 
with glass wool. The lower hole of the Fresenius tower is fitted with a rubber 
stopper, carrying a Geissler three-way stopcock (D). To one long tube of this 
stopcock are connected, by means of a second Geissler three-way stopeock, (EF), 
two manometers (IF and G), the former to indicate the degree of exhaustion and 
the latter the return to normal pressure. A Geryk air pump and a gasholder 
(IX) are connected by means of an ordinary three-way stopeock to the other arm 
of the Geissler three-way stopcock (D). Between the three-way stopcock and the 
air pump are imposed two drying tubes, the first containing calcium chloride 
* J, Am, Chem, Soc. (1906), 28, 223. 
* The apparatus worked splendidly with non-coking coals, but a coal retort of 
this size might be troublesome with coking coals. | 
The only advantage which a Geissler three-way stopcock has over an ordinary 
T-tube is to facilitate the discovery of leaks in the apparatus. 
