Discriminating Fossils – the crystal models belonging to the Watt family, c.1800

by Jane Insley (curator) and Valerie McCathern (volunteer; this project was Valerie’s fault, so she is co-author!)

Science Museum, London.

Image One: Watt workshop

The Science Museum has recently opened a new permanent exhibition about the 18th century steam pioneer James Watt. This had two main aims – one, the redisplay in public view of the garret workshop James Watt set up in Heathfield, his retirement home, and the other, to make more sense of the huge steam engines in the front hall. This is the story of one of the items from the workshop.

In the beginning was a box; a nondescript instrument box, sitting at one end of a bench in the James Watt Garrett workshop. Described in the inventory as ‘Oak box containing 209 Boxwood prisms in 24 cardboard boxes’

It had sat there for many years, twice been removed, packed up, transferred to a new home with the rest of the workshop, and then carefully replaced in position. There was no particular clue to the identity of the prisms.  Nearby there was another box containing 118 Boxwood prisms in 6 cardboard boxes.

Now James Watt died in 1819, and the story goes that the workshop was locked up, and no-one disturbed it until much later in the century. The iconic status that was growing around Watt as a national treasure had reached sufficient intensity by the 1860s that the existence of the workshop – the great man’s retreat and store room for things that covered his entire life – was a matter of public record, and concern for its preservation. In the meantime, the house was lived in by tenants, one of whom took many photographs of the inside of the house, available on-line.  The album included this photo of the word “Crystal” spelled out in little wooden shapes which indicates that the models had certainly been played with, probably by the little boy, and that then they were thought to represent crystal shapes. This blog post is a case study of how we tried to find out what an object is when all we knew was its provenance – owned by a famous man, what his interests were, and any information coming from an in-depth study of the object itself – in this instance, over 300 little crystal shapes. The work involved a tedious scanning job, comparison of two documents in an unfamiliar language, and the devotion of a volunteer; this provided core information for hunts in three other libraries, for documents scanned and unscanned, and the matching of objects to images. The result? The name of the maker, and date of production narrowed down to a 5 year period.

One day in 2010, I was showing my volunteer Valerie around the workshop. I was working as a content researcher for the redisplay project; Valerie had just finished a project to scan hundreds of record photographs in the collection, and was looking for something else to do. We opened the box [Image Two]. The prisms sat there in trays, made of cardboard and covered with blue sugar paper. They appeared to be completely randomly arranged. Valerie was much taken by them, as they were so nicely made. Carved from wood, they nevertheless were very crisply made, with sharply defined edges. Some of them had been marked in ink, with numbers, letters or a combination of the two. So she turned to me with a bright happy smile, and said in the way only volunteers can get away with, “You know Jane, you really ought to find out what these are, and how old they are, and who made them!”

Image Two: Picture of box, open

Urgghhh. My knowledge of 18th century mineralogy was non-existent, but I spoke to myself severely and reminded myself that when the going gets tough, museum curators go back to the objects, and so that is what we did. I graciously agreed to let Valerie scan the models. The trays were given identifying labels, and the models were scanned with the mark if any facing down, and with a scale giving length and breadth. This way, we preserved the arrangement of the models within the larger box, although, as we have seen, it was already truly compromised.

We then sorted the images by number, which showed that there were several runs of numbering within the collection. There were four “ones”, several marked with single letters, several marked with two letters, and a handful not marked with anything at all. The problem was then to try and find a suitably dated mineralogical work with plates of drawings of models, to which we could compare the numbering systems and shapes we had. From the bibliographic point of view, the two most likely candidates were Romè de l’Isle and the Abbe Haüy. Watt was known to have owned a book by de l’Isle, and the museum held a set matching the book in the Experimental Chemistry collection. And although we also held a set of Haüy models, they did not resemble the Watt collection in any way.

However, within the microfilm holdings of the Science Museum Library is a copy of the Watt papers held in the Birmingham Central Library. And in the Miscellaneous papers JWP 3/34, there was a “Manuscript description in German of a collection of c200 Werner’s crystals, nd[c1820] “. So we printed off a copy. Neither of us speaks German.

Now the crystals were wanted for filming for a multi-media presentation for the exhibition, so while Valerie worked on that, I went back to Google and the library catalogue and tried to find out a bit more about Werner.  He was a distinguished head of the Freiberg Academy, responsible for training in mines, but his written works were all in German, except for one, which described a classification system for minerals but alas did not include illustrations of shapes. So the next port of call was the library of the Natural History Museum next door, where they did have Werner’s works, and conversation with a very helpful librarian Camille Reeki threw up some other candidates in their catalogue. The key to the whole thing turned out to be a Google search using words from the title page of the manuscript, which drew out a book from the Special Collections by Carl Immanuel Löscher.

Löscher had been Werner’s assistant, and published “Beschreibung der Kristallisationen” in 1796, with a second edition in 1801. The second edition was not in the NHM Library, but had been scanned by Google; the first edition though had a beautiful set of plates in the back, which we were allowed to photograph, (but not photocopy – how odd is that?) and we went back to the office to compare them with the manuscript and the models.

Image Three: Valerie matching crystals to images

Valerie then set about matching the models to the plates. Working with an A3 size printout, she found that the match was very very VERY close – of four plates, only a few shapes here and there were not matched with a model. And it also included subsidiary number runs which matched the lettering systems as well.  It HAD to be right.  And when the remaining models were put in order, it became the main part of a single second set, with numbers from 1 to 200

Image Four: Set of four plates

Now we downloaded the second edition of Losher’s book from the web, but unfortunately the plates had not been opened out before scanning, so there were only tantalising glimpses of the shapes in the book. However, the first few at least matched some of the engravings in the second edition plates.  We needed a copy of the plates from the second edition. Fortunately for me, Wendy Crawforth of the Geological Society Library suggested that there were copies in both the British Library and the University library, Cambridge. A friend of mine was a regular visitor to Cambridge and within a couple of days she had ordered photocopies to be sent to me.

In the meantime, Valerie was comparing the texts of the manuscript and the printed edition. Because of our linguistic inadequacies, she hit upon the idea of highlighting the words that were the same from copies of the two texts, and found that with very small additions to the printed version, they were very similar. So it now appeared that one set of crystal models matched the first edition of Löscher, and the manuscript held in the Watt papers matched the second fairly closely. It was not clear which came first, but the suspicion was that the manuscript was intermediate in date between the two printed versions.

The photocopies arrived from Cambridge. Valerie set to again with the remaining 200 (odd) and concentrated on matching shapes. She found that the numbering systems jumped about, but when models were placed on engravings that matched, the jumps in the numbers on the models matched the equivalent number in the manuscript.  So we had a set of models matching Löscher ‘s book of 1796, and another set that match a manuscript which appears to be in between the dates of the two editions. It was not clear who had written or copied the manuscript. The writing was fluidly Germanic in style, but the only other clue was on the reverse of the back page, where “Description of a Collection of Werner’s Chrystals” was scribbled on the back. The handwriting was close to that of James Watt Junior – James Watt’s elder son. So which Watt owned the crystals?

Image Five: Set of 6 plates

James senior was known to have collected mineral specimens during his time installing steam engines, particularly in mining areas such as Cornwall. In the 1760s, when he first married, he had been involved with a pottery in Glasgow, and as an entrepreneur would have been looking out for novel sources of colours and other materials for his manufactures.

The younger son Gregory was a brilliant boy, whose talent was marred by ill-health, and eventually he would succumb to TB as his younger sister Jessie had some ten years before. Between the dates of the two versions of the printed book, Gregory appears to be the only member of the Watt family to travel in Europe, with part of his Grand Tour from Aug 1801 to the beginning of 1803 being specifically devoted to collecting mineral samples in northern Italy. This followed an earlier training in Cornwall where Gregory had been sent partly for health reasons, but with instructions from his father to William Murdoch who was keeping an eye on him to make sure that Gregory learned about the different types of minerals while he was there. And Gregory had an ongoing interest in the formation of basalt as it melts and crystallises which he wrote up for the Royal Society in a paper published in 1804, just before he died.

Than we come to James jnr. James ran a bit wild when he was young – didn’t get on terribly well with his step mother, and took a while to settle down before joining his father in the steam engine business. However, part of his education included being sent round Europe on an extended study tour in the 1780s, and I gather from a member of staff of the Freiberg Academy I met at another conference that James was enrolled as a student there, just as Werner was devising his theories, and Löscher became his second in command. To date, a published letter from James senior to James Junior gives the strongest indication we have that the models were James Junior’s. It goes –

Birmingham, May 7th 1787

I am glad to hear you are making such progress in your mathematics. As to Mr Werner’s classification of minerals, though it may be proper for you to know it, yet I cannot conceive of it to be of much consequence to you, as the chemical analysis is the only sure method of discriminating fossils, and is abundantly more simple than any other and must be reverted to in the long run, at least for all fossils that admit of it. That of the external characteristics has a few partisans in France, but none that I know of here.”

James does appear to have taken his father’s words on board, but developed an extensive collection of mineral samples of his own – Gregory sent three trunks full back from Naples on his tour – kept up with mineralogical literature in English, French and German, and in 1794 was ready to defend “his old master” Werner  from a bout of plagiarism. There is more detail to be unravelled here, as the models are later in date – ideally, we would like to find a receipt, or a letter ordering the books and the models, but the search continues.

Leave a Reply

Your email address will not be published. Required fields are marked *