More speculation on the source of sulfur in the Mary Rose, a plea to mitigate the influence of metrics and spontaneous combustion ignites debate.

Sulfur in the Solent

From Michael Baldwin

I have no doubt that Robert Close is correct in suggesting that gunpowder is behind the 2 tonnes of sulfur found on the Mary Rose. However, it would probably not have been as gunpowder, but as sulfur itself. 

At the time of the Mary Rose, there were two kinds of gunpowder: serpentine and corned. Serpentine was a mixture of sulfur, saltpetre and charcoal powders. Corned powder was a granular material made by pressing moist serpentine, granulating the resultant blocks and drying the product. Corned powder is much more powerful. 

Serpentine was used in early cannon while corned powder was used in hand guns. Writing in 1560, Peter Whitehorn in Certain waies for the ordering of souldiers in battleray noted: ‘If serpentine be used in hand guns, it would scant be able to drive their pellets a quoit’s cast from their mouths; and if corned powder be used in pieces of ordnance, without great discretion, it would break or mar them.’ Metallurgy had not kept pace with gunpowder technology.

It is likely, therefore, for safety reasons, that the three ingredients of gunpowder were held on board ship separately, and only incorporated into serpentine when needed.

As far as Whitehorn’s comments on the power of serpentine are concerned, as a teenager I found it difficult to make bangers with it, but it worked extremely well in a home-made gun constructed from drawn steel tubing.

M K Baldwin CChem FRSC 

Kent, UK

Taking the measure of metrics

From Anon

The news ‘Metrics role in assessing research to be reviewed’ is to be welcomed. 

Critics frequently point out some of the more extreme types of behaviour that occur when metrics are used to assess research, and the problems of metrics themselves. While I tend to agree with most of the criticisms, perhaps they distract from the fundamental problem: misuse of the data, or misunderstanding of how it can be used. Statistical data based on relatively weak correlations is only valid when a sufficiently large sample size is studied. For example, I am sure I could find a subset of the population who smoke, but live longer than average, but all studies using large samples show unequivocally that smoking will shorten your life. 

As scientists, we should be able to accept that there is only a fairly weak correlation between, say, citation volume and research impact for the majority of scientific research. This means that you can only use the data when the sample size is large enough to smooth out errors in both directions, and that it is unfair to make a comparison of only a few dozen papers such as happens on a recruitment or grant interview committee. 

This review, if it chooses to do so, should be able to assess and comment on what sample sizes will generate very reliable conclusions. The sample size of every paper published by a department of chemistry in a five-year period seems to me large enough for there to be quite a good correlation between various metrics and research quality. It is simple to measure the total volume of citations over a given period or an h-index for a whole department, or the volume of papers achieving more than a certain number of citations. 

All of this uses a large sample size, in terms of both papers and research fields. It also removes the false link that is made between research quality and citation volume of individual papers or an individual’s body of work. I would even argue this rapid measurement, combined with narrative reports of the department and maybe a small selection of the best papers to be reviewed, could replace the huge paper-reading exercise of the UK’s REF (research excellence framework). 

Please let us apply the same rigour in analysing the quality of the data being produced in research assessment as we do in conducting our own science.

Name and address withheld

When the wind blows

From Mark Foreman

In David Jones’s article ‘How to survive a nuclear bomb’, Jones explained that he planned to survive a nuclear weapon attack by entering a metal boiler and plugging up any holes to prevent the entry of radioactivity. 

I would like to point out that a typical boiler is unlikely to have sufficiently thick steel to attenuate the ? rays from radioactive fallout spread on the ground. If we assume that the isotope signature of the fallout is the same as that from the first Soviet atom bomb, which contaminated a nearby village, then in the first days 97Nb, 132I, 134I, 135I, 140Ba, 140La and 142La will be the main ? emitters.

While iodine can pose an internal threat, only that which enters the shelter could do so. I would be more concerned about the external ? hazard. I would expect most of the radioactivity to be spread on horizontal surfaces nearby.

In one way, he is right: you could use a car. One shelter design I have seen is to park a car over a trench and then fill the car with sacks of earth to improve the shielding (the thick earth walls will provide more shielding than a boiler). 

I would be inclined to drive the car to a motor mechanic’s workshop and use the inspection pit there as a readymade trench. An alternative would be to park an earth-filled bus before climbing down a manhole under the bus.

While nuclear warfare might captivate the mind, I would also like to point out that there is a lot more to nuclear science than gloom, doom and bombs.

Mark Foreman CChem MRSC

Chalmers University of Technology, Sweden

Precious platinum 

From Philip Leigh

In my previous employment at the (now closed) Sanofi site at Dagenham, UK, I developed a simple procedure for recovering platinum from waste organoplatinum oncology products. The method was based on the gravimetric analysis of platinum described in Vogel’s textbook of quantitative chemical analysis. Over two years the company received about £350,000 from selling the recovered metal. 

Although the methods developed at Dagenham were applied to waste oxaliplatin products they should be applicable to other organoplatinum products based on cisplatin and carboplatin. If they have not already done so, I would hope that other manufacturers of these oncology products would start to recover their own waste rather than sending it for incineration and subsequent landfill. This will not only save them money but will save a valuable and finite resource.

P Leigh CChem MRSC

Chelmsford, UK

Burning questions

From Walter Cuthbert

In response to David Jones’ theories on spontaneous combustion, I offer two examples of the phenomenon from my experiences in a large chemical factory in 1964. 

In one production area, plasticisers were made from coal-tar cresols and xylenols by reaction with phosphorus oxychloride using aluminium chloride as catalyst. The unit featured a continuous distillation train to remove the excess phenols and catalyst residues. After the startup trial, the main distillation column was allowed to cool and opened by the commissioning team for inspection. A process operator reported to me that ‘a tongue of fire’ had spat at him as he passed by the open manhole. I reported this and we were comprehensively ridiculed. The column was resealed and the low pressure distillation restarted but the temperatures inside the main column went beyond control even with the heating system switched off. The factory manager asked me if a fire was possible and I drew his attention to my report in the shift log. Removing the lagging revealed that the stainless steel column was glowing – the unit was on fire.

On another occasion, a tanker delivering phosphorus oxychloride sprang a leak on the connection to a bulk storage tank. The spillage covered the wide concrete roadway in a thin film. The works emergency crew started to hose the large area down with water. To our astonishment, a pale blue flame flickered across the surface of the spillage. I attributed this to traces of phosphine. Again, my report was treated with hilarity. 

I understood the scepticism; the classic hazard triangle is that air, combustible material and a source of ignition are required for a fire or explosion. It didn’t help my credibility that we were making fire-resistant plasticisers.

Walter Cuthbert CChem FRSC

Cheshire, UK

From John Twibell

David Jones seems to think that spontaneous human combustion (SHC) is a ‘bigger puzzle’. As a trained chemist who spent much of his career as a forensic scientist and fire investigator, I wish to point out that SHC is an enduring myth that is promulgated in some (usually non-scientific) circles, where facts have been changed and stories rewritten to add further mystery. 

Bodies do not spontaneously combust; they or their clothing become ignited in a conventional way, but in some circumstances, they burn in an unusual manner. The unusual features of such cases are all explainable by modern fire science. I investigated a ‘typical’ case a few years ago where an old man had died seated in an armchair and the major part of his body and chair had burnt away to ashes leaving only the extremities, his head and lower legs, largely unburnt. Items close by were undamaged. 

May I refer readers to an article I wrote about this case and a review of SHC in the Fire and Arson Investigator, January 2012, p53. This was published in response to the shock felt by the fire investigation community at a coroner’s decision in Ireland in September 2011, to record ‘spontaneous combustion’ as a cause of death. 

Jones’ last point about ‘seeing this happen to a joint of meat’ has been covered many times in recent years with demonstrations of pig carcasses being made to burn in this unusual manner. In a BBC TV episode of QED broadcast 26 August 1998, viewers were shown how the wick effect of clothing can produce a relatively small fire that gently dries out flesh and renders down the fat to give a slow burning fire that steadily consumes the body.

J Twibell CChem MRSC

Devonshire, UK