Report of the Purfleet Committee to the Royal Society
Printed in the Royal Society, Philosophical Transactions …, LXIII (1773–74), 42–7; AD (draft): Library of Congress.3
The Purfleet committee had its first meeting, presumably organizational, on July 18. On August 7 it visited the arsenal, which Franklin had already inspected in May, and then held four meetings in London between August 12 and 21.4 At the first one the members unanimously agreed on recommending metal lightning rods for the five buildings that constituted the magazine, and on where they should be placed and how fastened, but disagreed on what kind of rod to use.5 Four of the five men present advocated pointed rods, for reasons that Franklin had made familiar. For almost a decade, however, Benjamin Wilson had been harping on the danger of such rods and the greater efficacy of those that ended in blunt knobs,6 and he had no intention of changing his mind. On the 17th the committee considered his views, and the next day rejected them. A principal reason was two papers that Franklin read to his colleagues. One was the experiments and arguments based upon them that are printed above, under August 18. The other was a report of experimental work by Henly that buttressed Franklin’s position.7 Wilson remained totally unconvinced. At the final meeting on August 21, when the committee’s report was ready to be signed, he asked Franklin to give up the idea of pointed conductors. “No,” came the answer, “I will never give it up.”8 Wilson thereupon wrote a formal dissent, which went with the report to the Council of the Royal Society and thence to the Board of Ordnance.9
Such controversy over the form of the conductors, although it seems at first glance like that between Big-endians and Little-endians, in fact went to the heart of the question of how lightning behaves. Wilson’s position can be briefly summarized.1 He agreed that points conduct more efficiently than knobs, and hence attract more readily the electricity in an approaching cloud. For that very reason, he insisted, points should not be used: they tended to “solicit” a lightning stroke that otherwise might not occur. No one knew, furthermore, how great a charge might build up in a thunderstorm; points, because they were more effective than knobs, could draw a greater stroke if one did occur.2 On the same ground he challenged the committee’s recommendation that the rods be raised high above the rooftrees: the taller the rods, the greater the likelihood of their coming within the striking distance; hence the shorter the better. Lastly he claimed that the rods his colleagues were recommending were far too small; the larger the conductor, according to him, the greater the chance of its conveying the charge safely. He did not offer, as Franklin did, experimental observations of his own to buttress any of his arguments; and his colleagues found mere assertion unconvincing.3
August 21, 1772.
The Society being consulted by the Board of Ordnance, on the propriety of fixing conductors for securing the powder magazines at Purfleet from lightning, and having thereupon done us the honour of appointing us a committee, to consider the same, and report our opinion; we have accordingly visited those buildings, and examined, with care and attention, their situation, construction, and circumstances, which we find as follows.
They are five in number, each about 160 feet long, and about 52 feet wide, built of brick, arched under the roof, which in one of them is slated, with a coping of lead 22 inches wide on the ridge from end to end; and the others, as we were informed, are soon to be covered in the same manner. They stand parallel to each other at about 57 feet distance, and are founded on a chalk rock, about 100 feet from the river, which rises in high tides within a few inches of the level of the ground, its brackish water also soaking through to the wells that are dug near to the buildings.
The barrels of powder, when the magazines are full, lie piled on each other up to the spring of the arches; and there are four copper hoops on each barrel, which, with a number of perpendicular iron bars, (that came down through the arches, to support a long grooved piece of timber, wherein the crane was usually moved and guided to any part where it was wanted) formed broken conductors within the building, the more dangerous from their being incompleat, as the explosion from hoop to hoop, in the passage of lightning drawn down through the bars among the barrels, might easily happen to fire the powder contained in them. But the workmen were removing all those iron bars (by the advice of some members of this Society, who had been previously consulted4); a measure we very much approve of.
On an elevated ground, nearly equal in height with the tops of the magazines, and 150 yards from them, is the house where the Board usually meet. It is a lofty building, with a pointed hip-roof, the copings of lead down to the gutters, from which leaden pipes descend at each end of the building into the water of wells of 40 feet deep, for the purpose of conveying water forced up by engines to a cistern in the roof. There is also a proof-house,5 adjoining to the end of one of the magazines, and a clock-house, at the distance of [blank in text] feet from them, which has a weathercock on an iron spindle, and, probably, some incompleat conductors within, such as the wire usually extending up from a clock to its hammer, the clock, pendulum, rod, &c.
The blowing up of a magazine of gun-powder by lightning, within a few years past, at Brescia in Italy, which demolished a considerable part of the town, with the loss of many lives,6 does, in our opinion, strongly urge the propriety of guarding such magazines from that kind of danger; and since it is now well known, from many observations, that metals have the property of conducting lightning, and a method has been discovered of using that property for the security of buildings, by so disposing and fixing iron rods, as to receive, and convey away, such lightning as might otherwise have damaged them; which method has been practised near twenty years in many places, and attended with success, in all the instances that have come to our knowledge,7 we cannot, therefore, but think it adviseable to provide conductors of that kind for the magazines in question.
In common cases, it has been judged sufficient; if the lower part of the conductor were sunk three or four feet into the ground, till it came to moist earth; but this being a case of the greatest importance, we are of opinion that greater precaution should be taken. Therefore, we would advise, that, at each end of each magazine, a well should be dug in or through the chalk, so deep as to have in it at least four feet of standing water. From the bottom of this water should arise a piece of leaden pipe, to or near the surface of the ground, where it should be strongly joined to the end of an upright iron bar, an inch and half diameter, fastened to the wall by leaden straps, and extending ten feet above the ridge of the building, tapering from the ridge upwards to a sharp point, the upper 12 inches of copper, the iron to be painted.
We mention lead for the under-ground part of the conductor, as less liable to rust in water and moist places; in the form of a pipe, as giving greater stiffness for the substance; and iron for the part above-ground, as stronger, and less likely to be cut away. The pieces, of which the bar may be composed, should be screwed strongly into each other, by a close joint, with a thin plate of lead between the shoulders, to make the joining or continuation of the metal more perfect. Each rod, in passing above the ridge, should be strongly and closely connected by iron or lead, or both, with the leaden coping of the roof, whereby a communication of metal will be made between the two bars of each building, for a more free and easy conducting of the lightning into the earth.
We also advise, in consideration of the great length of the buildings, that two wells, of the same depth with the others, should be dug within twelve feet of the doors of the two outside magazines; that is to say, one of them on the north side of the north building, the other on the south side of the south building; from the bottom of which wells, similar conductors should be carried up to the eaves, there joining well with a plate of lead, extending on the roof up to the leaden coping of the ridge, the said plate of lead being of equal substance with that of the coping.
We are further of opinion, that it will be right to form a communication of lead from the top of the chimney of the proof-house to the lead on its ridge, and thence to the lead on the ridge of the corridor, and thence to the iron conductor of the adjacent end of the magazine; and also to fix a conductor from the bottom of the weather-cock spindle of the clock-house, down on the outside of that building, into the moist earth.8
As to the board-house, we think it already well furnished with conductors, by the several leaden communications above-mentioned, from the point of the roof down into the water, and that, by its height and proximity, it may be some security to the buildings9 below it; we therefore propose no other conductor for that building, and only advise erecting a pointed iron rod on the summit, similar to those before described, and communicating with those conductors.1
To these directions we would add a caution, that, in all future alterations or repairs of the buildings, special care be taken that the metalline communications be not cut off or removed.
It remains that we express our acknowledgements to Sir Charles Frederick, Surveyor-general of the Ordnance, for the obliging attention with which he entertained and accomodated us on the day of our enquiry. With very great respect, we are, Gentlemen, Your most obedient, humble servants,
3. The verbal differences in the draft are, with a few exceptions mentioned below, too minor to be worth notice.
4. See above, the headnote on BF to Dawson, May 29; Robertson to BF, July 14; Cavendish to BF, Aug. 4; and the committee’s minutes under Aug. 12.
5. See ibid. and Benjamin Wilson, Observations upon Lightning … (London, 1773), pp. 15–16. Edward Hussey Delaval (above, VIII, 359–60) was appointed to the committee but was unable to attend; he subsequently supported Wilson’s position: Observations, pp. 19–35.
6. See his paper in Phil. Trans., LIV (1764), 247–53.
7. BF mentioned, in the committee’s letter below under Dec. 10, that Henly sent him the report (which is now lost) and that it demonstrated the ability of a point to draw electricity from a cloud when twelve times more distant from it than a knob. In Further Observations upon Lightning … (London, 1774), pp. 2–3, Wilson identified the experiment most damaging to his theories as the fifth in a series that Henly had performed the previous April and later published in Phil. Trans., LXIV (1774), 133–52. In his copy of this pamphlet, acquired by BF and now in the Yale University Library, Henly wrote, “the 5th fully proves that the point would not … invite a stroke upon itself at twelve times the distance of the knob. …” He presumably meant that the experiment, in which the twelve-to-one ratio does not appear, provided corroboration for another in which it does. The latter was undoubtedly that which he described in the volume of the Phil. Trans. just cited, pp. 141–2, 409–10.
8. Wilson, Further Observations, p. iv.
9. The dissent was published in Phil. Trans. immediately after this report.
1. He expounded it at length in the two pamphlets cited above.
2. Wilson adduced as evidence Henly’s experimental work that has just been discussed. It demonstrated, he mistakenly believed, that a point would be struck when twelve times as far from the cloud as a knob and also, apparently, that it would receive twelve times the charge. Observations, pp. 4–5, 15, 57. Henly’s marginal comments on this argument in his copy of Further Observations have been badly trimmed, and most of what is left reveals little. But one comment, facing the title page, sums up the controversy. Henly challenged Wilson by letter, to which he received no reply, to “prove by experiment, that placing a knob, and a point, at an equal distance from a charged body, the point would be struck with the smallest force: or that making the charge exactly the same, the point, could be struck with it, at the greatest distance. …” This is substantially what BF had said a decade earlier: above, X, 51–2; see also Henly in Phil. Trans., LXIV (1774), 138, 152.
3. Modern investigation has established that the practical difference between pointed and blunt rods is negligible, but has settled the other questions at issue in BF’s favor. See Sir Basil F. J. Schonland, “The Work of Benjamin Franklin on Thunderstorms and the Development of the Lightning Rod,” Franklin Institute Jour., CCLIII (1952), 375–92, and The Flight of Thunderbolts (2nd ed., Oxford, 1964), chaps. 2 and 6.
4. Wilson and BF; see the headnote on BF to Dawson above, May 29.
5. For testing the powder.
6. See the headnote just cited. In his draft BF added, then deleted, a reference to the similar explosion in Jamaica that is mentioned in Small to BF above, April 13.
7. BF heard, a few months later, of an example to the contrary, but it proved to be unfounded: to Bache below, Oct. 7. Some years earlier BF had recommended protecting a powder magazine with a single rod placed near but not on the building: above, X, 183–4.
8. The draft here adds a sentence that was later deleted, either by BF or some one else: “For tho’ the Proof-house and Clock-house have not generally any Gunpowder in them, yet as Lightning sometimes rends large Stones out of Buildings, and throws them about with Violence, so as to penetrate and damage other adjacent Buildings, to avoid any Danger, however unlikely to happen, from such Collision, we think it adviseable to secure them.”
9. So in the draft; the printed text reads “building,” which must be a misprint.
1. Most of the committee thought this rod redundant, but apparently agreed to it to meet Wilson’s charge that they were being inconsistent; see his Observations, pp. 5, 15–17.