I recently came by a set of W.F. Stanley railway curves made of vulcanite (also known as ebonite or hard rubber), included almost as an afterthought to an ebay listing for an Elliott Bros. surveyor’s level that I ended up acquiring as collateral (and which turned out to be quite interesting in its own right, but that’s something for another time). From the poor photos it was not possible to determine if the set was complete or damaged, nor the number or size of the pieces, but vulcanite curves rarely turn up for sale in any condition so I bought it essentially sight unseen.
What became clear when it arrived is that there were an awful lot of curves in the box, certainly somewhere in the region of the number in my full Thornton set of 100. The larger curves were similarly 18 inches in length, but without the staggered lengths seen in the Thornton set (intended to equalise the heights of the various curves). However, unlike all my other sets, they were not stacked vertically, but placed flat in their mahogany box in bundles, two of which were wrapped in paper with handwritten numbers on.
I decided to take a systematic approach and remove one curve at a time, cleaning off any dust (vulcanite is notoriously dusty due to its tendency to generate static electricity) and making a note of the radius number before returning them all to the case vertically to match my other sets. By the time they were all laid out on the floor and I had counted them up, I was surprised to reach a total of 108 – so much so that I counted them again, and then counted my list to double check. Moreover, the radius numbers were oddly erratic, as indicated by the handwritten paper wraps, which is why I had initially assumed the set would be missing curves. Instead, it seemed to have more than it should.
When I began to return them to the box, I was greeted with another surprise: there were simply too many curves to fit in a vertical orientation, nor was it possible to fit the smaller ones down the side in the manner of the Thornton set. Even 100 of the vulcanite curves could not be accommodated in this way. This, along with the very neat arrangement of the three bottom sets of curves (including the two paper-wrapped bundles) suggested that the curves had been deliberately stored horizontally, possibly to avoid warping. It led me to wonder if the paper wraps might have been original to the set as supplied, perhaps as a custom order. Although deteriorated, the wraps were neatly folded from good quality envelopes, using the gummed edge as adhesive, a technique that I have encountered before in sets of both stencils and rulers. Unfortunately on this occasion there was no watermark to help date the paper.
Turning to the set’s mahogany case for clues, the comb-jointed corners suggested a date after the move to Stanley’s new factory in 1876, but the lack of hooks and style of lock was more typical of their earlier joinery work.
There was no label inside the lid, but this may have been lost due to the damage and subsequent repair work that had clearly taken place at some point in the past. Finally, an upside-down pencil inscription inside the case appeared to read “K 11 98”. This could be a date – November 1898 – or perhaps item K1198 from Stanley’s 1905 K edition catalogue (which I have not yet been able to check, although the catalogues to either side suggest this is unlikely). Equally, it could refer to the 11-inch and 98-inch radius curves (both of which are included in the set), or perhaps was just an arbitrary code used during the assembly of the box.
The curves themselves are also difficult to date. As a whole, they appear unused, none of them showing any sign of damage or significant wear. They definitely predate the early 20th century change to Stanley’s “Trade Mark” logo, but the simple white-filled “Stanley London” inscription could go back as far as the introduction of vulcanite to Stanley’s range (see, for example, this set of 19th century chain scales). A little research showed that this occurred during the 1860s.
Stanley’s catalogue dated 1865 includes a boxed set of 50 railway curves in pearwood at £1 7 0 (27 shillings) and in cardboard at 18 shillings. while vulcanite is not listed as an option, it was clearly available by this date as a note below the curves section of the catalogue reads “Vulcanite and Metal Curves”, presumably price on application. That cost was indeed the issue here is confirmed by Stanley himself in the 1868 edition of his Mathematical Drawing Instruments:
“The vulcanite mentioned for set squares makes very excellent radii curves for temperate climates, but the cost of the material prohibits their general use.”
Not one to miss an opportunity to claim his pioneering use of vulcanite – in much the same way as he had patented the manufacture of drawing instruments from aluminium in 1866 – Stanley waxes lyrical about the new wonder material:
“The author, after several experiments in seeking some suitable materials for plain set squares, has discovered that Goodyear’s vulcanite, which is a patented preparation of India rubber, of which the best qualities are manufactured in North America, possesses all the qualities desirable for set squares to be used in temperate climates. This material is considerably harder and tougher than any kind of wood; it is impervious to moisture, consequently it may be kept clean, if required, by washing, and it will not warp or get out of truth under any ordinary circumstances.”
This implies that Stanley was buying Goodyear‘s vulcanite at the time, so would have been paying a premium due to both import costs and patent royalties. To give an idea of its relative cost, set squares made of “Stanley’s vulcanite” were listed in his 1865 catalogue at 3s 6d for the largest 10-inch model, a full two shillings more than the pearwood equivalent at 1s 6d. Likewise, the price list at the back of his 1868 book includes vulcanite set squares, but not railway curves. Their first appearance comes in Stanley’s price list for 1869, where a set of fifty in vulcanite would have set you back 55 shillings. This was still just over twice the cost of pearwood, which remained at 27 shillings for a box of 50, and three times the price of cardboard at 18 shillings.
Even at this level, Stanley may have struggled to make a profit, as by 1873 the cost of a vulcanite set had actually risen to 60 shillings with pearwood unaltered at 27 shillings. The 1873 list was also the first to offer a set of 100 in vulcanite, at a whopping five pounds (100 shillings). There was no change in the prices at the back of the 1878 and 1888 editions of Stanley’s book, but by 1900 both pearwood and cardboard curves had risen to 32s and 20s respectively, while vulcanite stayed at 60s for a set of 50. This may be the first indication that the cost of vulcanite was beginning to drop relative to other materials.
The prices remained this way in the 1909 J edition catalogue. However, they were now joined by the new kid on the block, transparent celluloid, at 70s for 50 and 120s for 100, more expensive even than vulcanite. By 1919 the cost of all Stanley’s curves had doubled in proportion to those of 1909, with celluloid still the most expensive option, but by 1927 the price of celluloid had dropped to match that of vulcanite. This parity remained the case in 1931, when a set of 100 curves in either vulcanite or celluloid cost £9 15s.
In spite of this competition from celluloid, the 1925 edition of Stanley’s book (revised by his long-time collaborator and successor in business, Henry Thomas Tallack) continued to make the case for the material:
“Vulcanite makes also very excellent curves. Transparent celluloid is in favour owing to its great convenience, but it is not reliable in the tropics.”
Perhaps for this reason, Stanley’s vulcanite railway curves even survived the Second World War, appearing in the firm’s first postwar catalogue and its related price lists until at least 1952. However, by 1958 vulcanite curves had disappeared from their range, leaving pearwood and “transparent plastic” as the only options. Whether the latter term meant celluloid or one of the newer acrylic plastics is unclear, but given the rapid improvements in synthetic polymers it is a wonder that vulcanite remained available as long as it did. As a material for scales, vulcanite had already been dropped from Stanley’s catalogues by 1909, its continued availability acknowledged only by a note that “Any scales can be made of vulcanite, but these are not recommended, as they expand greatly with heat.” By the 1931 “P” edition catalogue, even this note had disappeared.
These later catalogues also provide a final piece of the puzzle relating to my set of 108 curves. For the first time in 1924 the full content of Stanley’s largest set of 100 is listed:
1, 1 1/4, 1 1/2, 1 3/4, 2, 2 1/4, 2 1/2, 2 3/4, 3, 3 1/4, 3 1/2, 4, 4 1/2, 5, 5 1/2, 6, 6 1/2, 7, 7 1/2, 8, 8 1/2, 9, 9 1/2, 10, 10 1/2, 11, 11 1/2, 12, 12 1/2, 13, 13 1/2, 14, 14 1/2, 15, 15 1/2, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 34, 36, 38, 40, 42, 44, 46, 48, 50, 52, 54, 56, 58, 60, 63, 66, 69, 72, 75, 78, 81, 84, 88, 92, 96, 100, 104, 108, 112, 116, 120, 125, 130, 135, 140, 145, 150, 155, 160, 166, 172, 178, 184, 190, 200, 210, 220, 230 and 240
Oddly, this totals 101 curves, whereas the 25 and 50 curve sets both contain the advertised number. I assume this excess of one resulted from the addition of the 1-inch curve, not present in the two smaller cases. By comparison, Thornton’s equivalent set of 100, running from 1 1/2 to 240 inches, contains exactly 100 curves (I have counted the curves in my own Thornton set which bears this out, the smallest curve being the 1 1/2 inch). Going back to Stanley’s 1888 catalogue, the set of 100 runs from 1 1/2 to 240, which would seem to confirm that the 1-inch curve was added some time after this date.
Turning to my vulcanite set, it is missing four of the curves from the canonical 100 (namely 2 3/4, 3, 7 1/2 and 17), but contains an additional eleven supernumerary curves (33, 39, 41, 47, 62, 70, 80, 85, 90, 95 and 110). It will be noticed that these are all in the upper part of the register, where the gaps between consecutive curves become greater than one inch. The motivation appears to have been to fill some of the gaps in coverage, particularly where multiples of five or ten have been omitted from the standard set.
The set therefore seems to have been a custom order, which reinforces my impression that the numbered bundles of curves were supplied that way. This in turn supports the idea that vulcanite curves were intended to be kept flat in their box, not stacked vertically as with pearwood. I have seen examples of Stanley’s celluloid curves stacked vertically in compartments, but also the occasional small set of pearwood curves stored flat, with a specially-shaped block to keep them in place. What I would really like to see are more examples of Stanley’s vulcanite curves, which might help corroborate this theory.
In any case, the overall impeccable condition of these curves, now possibly approaching 150 years old, may be seen as vindication of Stanley’s high opinion of vulcanite. Were it not for the excessive cost of the material, it would surely have been more widely used.
