60 
Proceedings of the Royal Society of Edinburgh. [Sess. 
in this case the tree trunks had become lignite before the pressure of super- 
incumbent accumulating sediment had reached the stage at which oil 
formation could commence, but that once this stage had been reached the 
rest of the vegetable matter present had been converted into petroleum, the 
lignite remaining unchanged. 
This section occurs in what may be called the transition stage; the 
series above is entirely lignitic, while below, though there are some lignitic 
bands, the petroliferous phase becomes increasingly pronounced. 
Certain phenomena of natural filtration of petroleum and the weathering 
of oil-rocks must now be considered. 
In examining a series for evidence of petroleum it is natural to seek 
first the most porous beds, as these may be expected to contain the greatest 
quantity of liquid hydrocarbons. It does not follow, however, that this is 
the most sensible course to pursue except under certain conditions. If a 
coast or river section is being examined, and the strata are almost entirely 
argillaceous and impervious, any bed of sandstone, however thin, will 
be examined for traces of oil ; and if no porous bed be present, ironstone 
nodules and thin laminae or concretions of limestone will be examined. On 
breaking these a faint odour of inspissated petroleum may be noticed, or 
possibly bituminous stains or a film of dried-up petroleum on the joint faces 
may be observed. Thus where the mass of the strata is too impervious to 
contain distinct traces of an impregnation, any material slightly more 
porous may give evidence of having been impregnated by petroleum that 
has filtered slowly through the to all intents and purposes impervious 
surrounding rock. 
But where the strata to be examined have been fully exposed to 
weathering agencies, say in an outlier, it will be necessary to examine the 
less pervious beds. Porous sandstones which may have been oil-sands may 
have been so thoroughly weathered and lixiviated that not a trace of 
petroleum can be detected. Beds of clay and sandy clay, however, which 
are sufficiently impervious to have only been impregnated very slowly, 
yield their contents of liquid hydrocarbons equally slowly, and may retain 
traces of petroleum long after all more porous beds have lost every such in- 
dication. This phenomenon can often be studied to advantage when an 
oil-bearing series lies unconformably upon an older series that is not oil- 
bearing in a primary sense, and where the younger oil-bearing series has 
since been removed by denudation. The more porous beds of the older 
series may long have lost all traces of a former impregnation, but the less 
pervious beds may still be yielding a little oil and gas in seepages, or may 
