17G 



KNOWLEDGE 



[June 1, 1887. 



this question fully and in detail would be a very difficult 

 task, and is at present certainly impossible. But, speaking 

 generally, it may be said that the choice lies between two 

 theories, in greater or less vogue among geologists ; while, 

 so strongly does all the evidence bear on the side of one of 

 these, that it is very difficult to understand how any doubt 

 can remain as to its essential validit}'. 



Let it be premised that the oil occurs chiefly in what are 

 called pools, though they are not in the slightest degree like 

 pools — certain sands found at a greater or less depth beneath 

 the siu'face, and permeated with oil in such degree that in some 

 cases a cubic foot of oil is obtained from every ten or twelve 

 cubic feet of sand. Intermixed with the oil is a gas which 

 also is found collected in many places above the oil pools, in 

 a highly compressed state. This gas is variously constituted 

 in different wells. It always contains a large proportion of 

 marsh gas — the gas which rises from the ground above 

 places where animal or vegetable matter is undergoing pro- 

 cesses of decomposition. It also contains hydrogen, carbonic 

 acid gas (or what chemists now call carbon dioxide), ethyl 

 hydride, and several hydrocarbons. The petroleum itself 

 has been formed from the condensation of a portion of this 

 compound gas. The proportions of the several gases vary in 

 different wells. Thus, while marsh gas amounts in some 

 wells to nearly 90 per cent. (S9'C.3 is the highest amount 

 yet deduced fiom analysis, I believe), gas from the Cherry- 

 Tree well shows only GO-27 per cent, of marsh gas. Hydro- 

 gen ranges from 6-G to 13'5 per cent, in other wells, but in 

 gas from Cherry-Tree well amounts to 22 5 per cent. In 



like manner carbonic acid 



which has a range of from 



-j'i,';,th to TVi'otti per cent, in gas from other wells, runs 

 up in the Cherry-Tree gas to 2'28. Again, while there is 

 no trace either of oxygen or nitrogen in the gas from most 

 of the oil districts, we find in the gas from the Cherry-Tree 

 well /injth per cent, of oxygen and 7'32 per cent, of 

 nitrogen. Of ethyl hydride, which may be regarded as the 

 chief component of all the petroleum gases outside the 

 marsh gas and the permanent gases, we find the percentage 

 ranging between 4'39 and IS'39 in other wells, and amount- 

 ing to trSO in the Cherry-Tree gas. 1 have been somewhat 

 particular in noticing the exceptional nature of the gas from 

 the Cherry-Tree well, because it illustrates strikingly the 

 eflect which local peculiarities may have in modifying the 

 chemical constitution of the gas found in the oil districts. 

 The Cherry-Tree gas i-eaches the surface through water ; we 

 can under.^tand well, then, that it should contain a relatively 

 larger proportion of hj-drogen, oxygen, mtrogen, and car- 

 bonic acid gas ( with which the water is largely charged). 



It is woi'thy of notice, however, that the structure of the 

 sands in which the oil is chiefly found by no means affects 

 in similar degree, or even in any appreciable degree at all, 

 the character and chemical constitution of the oil. This is 

 the first, and perhaps most striking, piece of evidence show- 

 ing that the oil was not originally formed where it is now 

 found. For convenience I present the evidence on this 

 point, in company with other evidence about the oil regions 

 which is full of interest, though not bearing directly on the 

 question of the origin of petroleum. 



The sand from which the oil in the ^\.llegany (not Alle- 

 gheny) and Bradford districts is obtained consists of a grey- 

 black, dark-brown, or chocolate-brown coloured sand, singu- 

 larly uniform in texture, of about the coarseness of the 

 ordinary black sand of the New Jersey coast. The oil 

 obtained is dark amber-green in colour, and occasionally 

 black. The specific gravity is about half that of water, that 

 is about 50. The Bradford oil region has an extent of about 

 133 square miles, and up to Januai-y 1, 188-">, had produced 

 820,000 barrels per square mile, or 109,000,000 barrels in 

 all. The Allegany region has an area of 31 square miles, 



and to the same date had produced 15,000,000 barrels of 

 oil, or 419,000 per square mile. Now the sand of the cele- 

 brated Venango district consists of white, grey, and yellow 

 pebble rock — the pebbles being obviously water worn, and 

 as large as hazel-nuts, loosely cemented and bedded in fine 

 sand, which is by no means so regular as that of the Alle- 

 gany and Bi-adford regions. Hence there are dry tracts 

 (holes they are called, being no more holes than the rich 

 tracts of oil sands are pools) where the oil has not gathered. 

 The oO in the Venango district is generally green, but fre- 

 quently black, and sometimes amber. Its specific gravity 

 ranges from 30 to 51, but may be regarded as averaging 

 about 48. This district has an area of 65 square miles. The 

 total [)roduce to the beginning of 1885 amounted to 

 55,000,000 barrels, or 84G,t(00 barrels to the square mile. 

 The Butler district has an area of eighty-four square miles. 

 The sands resemble those in the Venango region, and the 

 oils obtained are similar in colour and specific gravity. To 

 January 1, 1885, the supply of oil from this region amounted 

 to 69,000,000 barrels, or to 821,000 barrels per square mile. 

 From the Beaver district, sixteen square miles in extent, 

 1,000,000 barrels had been obtained by the beginning of 

 1885, corresponding to 62,0(10 barrels per square mile. In 

 this region the sand rocks correspond with those called the 

 Pottsville conglomerate and the Berea grit in the sub- 

 carboniferous series. The oils obtained from the Butler 

 district are mostly amber-coloured. In the Warren region, 

 thirty-five square miles in extent, the sands vary greatly 

 in character. The total jiroduce to January 1885 was 

 12,000,000, or 343,000 barrels per square mile. 



When we notice in the Venango drillings coarse sand and 

 pebble, and pieces of jHilverised rock showing no traces of 

 petroleum, we cannot but feel how unlikely it is that such 

 a rock could ever have contained enough organic matter to 

 yield a cubic foot of oil to every ten or twelve cubic feet of 

 rock. Again, it is obvious that in such a rock substance 

 organic matter would decompose and waste. The organic 

 matter could not have found its way into those pebbly sands 

 while they were exposed to atmospheric influences, for in 

 that case the oil would have been volatilised and dissip.ated. 

 Nor could it have been accumulated when these sands were 

 in contact with the water ; for then the oil, being of much 

 lighter specific gravity, would have risen to the surface and 

 floated away as flist as it was formed. The oil must, then, 

 have been introduced into these sands long after they formed 

 part of any shores, or were near to the surface of the earth's 

 crust, whether that surface above these sands were under 

 water or not. 



But perhaps the strongest argument against the theory 

 that the oil was formed in situ resides in the very special 

 way in which the oil is distributed in layers at difl'erent 

 depths, but under the same place. Of cour.se, if each sandy 

 tract originally supplied its own organic material, animal 

 and vegetable, from whence, long after, oil was formed (the 

 region sinking to depths where the heat and temperature 

 were sufficient to decompose the organic material and 

 generate from it vast quantities of gas), we could not expect 

 any connection to be shown between the distribution of the 

 oil in one sand pool and the distribution in another lying 

 below it (at so great a depth as to indicate \evy long time- 

 intervals between the formation of the respective sands). 

 But a singular connection is actually recognised, the 

 meaning of which seems unmistakable. Wherever a tract 

 of oil sand overlies a lower rich oil sand the upper 

 sand is always poor. But if portions of a lower oil 

 sand are rich and other portions poor, then the upper 

 oil sand is rich where the lower is poor, and poor where 

 the lower is i-ich. Again, where an upper sand tract 

 extends farther than a lower one, the part overling the 



