ON THE ASSAY OF COAL FOE, CRUDE OIL, ETC. 
15 ^ 
that the relative proportions of oil, water, and gas in volatile matter vary so 
greatly, that of much volatile matter but little may prove to be oil, and of a 
small percentage of volatile matter nearly all may be oil; and, again, the dif¬ 
ficulty of applying a low and regular heat to such an arrangement of crucibles 
is so considerable, that I have met with difierences of 10 to 15 per cent, of 
coke from the same sample of broken and previously well-mixed coal. In 
one experiment four porcelain crucibles were placed in one large common 
crucible, the first porcelain crucible being placed near the bottom and the 
three others in the upper part, and all charged with the same prepared speci¬ 
men of coal. After ignition, the lower crucible ^fielded 9 per cent, more coko 
than the others; owing, mainly, doubtless to its receiving more heat than 
they—a high temperature, as is well known, causing decomposition of some 
of the hydrocarbons, of which oil and gas are formed, wnth separation of solid 
carbon. 
llejecting, then, the experiment of subjecting the coal, etc., to a preliminary 
coking operation, the material is at once submitted to distillation. This I 
usually perform in a piece of common iron gas-piping, a metre and a quarter 
long, and five to eight centimetres internal diameter. Two or three kilo¬ 
grammes of the coal, properly sampled, is then broken down, to the size of 
peas or hazel-nuts, avoiding the formation of dust. The fragments are well 
mixed together, and from one-half to one kilogramme weighed out for distil¬ 
lation. This quantity is poured into one end of the iron tube, a diaphragm 
of wire gauze, suspended by a long wire at about thirty centimetres from the 
other end of the tube, preventing the coal from falling through. The coal so 
placed occupies half a metre or more of the tube, leaving a free space of 
thirty or forty centimetres at either end of the tube. The region of the tube 
occupied by coal is now heated by a series of gas-jets until the lowermost part 
of the iron is just visibly red in a darkened room, the gas-jets being turned 
off during the moment of observation of the temperature. To heat the tube 
equally and not too highly by charcoal is difficult. A Hofmann’s gas-furnace, 
two-thirds of a metre long and having three rows of burners is perhaps the 
most convenient for the purpose ; it should be just fairly alight along its 
whole surface. These arrangements are, in my opinion, those most conve¬ 
nient for quantitative experiments ; if the object be merely the production of 
oil, a wider tube and a furnace of five rows of burners may be used. Tor 
experimental purposes, the iron tube and furnace should, I think, in nearly all 
cases be inclined downwards from the horizontal position at an angle of ten or 
fifteen degrees. Indeed, one office filled by the wire-gauze diaphragm already 
alluded to is to prevent the falling of the coal out of the region of the furnace 
wdien the iron tube is inclined as just stated. The advantage of inclining 
the tube is, that while nothing is lost which could possibly be formed in or ob¬ 
tained from a retort of the usual form, some of the paraffin and heavier 
vapours, which might under other circumstances remain about the source 
of heat, and become decomposed mainly into gas, at once flow oT downwards 
from the region of danger. The condensation of the vapours produced in the 
distillation is effected in a common bottle of one or two litres’ capacity, par¬ 
tially immersed in a vessel of cold water, a wet cloth covering the upper sur¬ 
face of the bottle and dipping into the water. This arrangement is usually 
sufficient even for the condensation of oils from coals containing a consider¬ 
able amount of moisture. The iron retort-tube is connected with the bottle 
by a hollow tin cone about a third of a metre long, the larger orifice fitting 
loosely over the end of the iron tube, the smaller passing through a cork' into 
the condensing bottle. The neck of a broken retort, commonjy used as an 
adaptor in laboratory distillations, may be used in place of the tin cone, but 
is more than usually liable to fracture. The tin is luted to the iron by plaster 
