Inventing Temperature Page 22

  By the time Josiah Wedgwood (1730-1795) started in earnest to experiment with pyrometry, he had established himself as the leading manufacturer of porcelains in Britain and indeed one of the most renowned in all of Europe, equaling the long-established masters in Meissen and Sèvres. Wedgwood was proudly called "Potter to Her Majesty" since 1765, a title that he received after fulfilling a high-profile order from Queen Charlotte for a complete tea set (Burton 1976, 49-51). Ever seeking to improve and perfect the many lines of products issuing from his Staffordshire works grandly named "Etruria," Wedgwood was very keen to achieve better knowledge and control of the temperatures in his kilns. In 1768 he wrote exasperatedly to his friend and business partner Thomas Bentley: "Every Vaze [sic] in the last Kiln were [sic] spoil'd! & that only by such a degree of variation in the fire as scarcely affected our creamcolour bisket at all."25

  For Wedgwood's purposes the standard mercury thermometer was of no help, since mercury was known to boil somewhere between 600°F and 700°F. In his preliminary studies, Wedgwood read an article by George Fordyce (1736-1802), chemist, physician at St. Thomas's Hospital in London, and a recognized authority on medical thermometry.26 Fordyce asserted that bodies began to be luminous in the dark at 700°F, and Wedgwood noted in his commonplace book: "If mercury boils (or nearly boils) at 600 how did Doctr. Fordyce measure 700 by this instrument." Later he had a chance to buttonhole Fordyce about this, and here is the outcome: "Upon asking the Dr. this question at the Royal societie's rooms in the spring of 1780 he told me he did not mean to be accurate as it was of no great

  23. Lafferty and Rowe (1994, 570) give 9980°F or 5530°C.

  24. For a late case of a relatively uncritical citation of Wedgwood's measurements, see Murray 1819, 1:527-529. Pouillet (1827-29, 1:317) used Wedgwood's data, even as he expressed dissatisfaction with the instrument.

  25. Wedgwood to Bentley, 30 August 1768, quoted in Burton 1976, 83; the full text of this letter can be found in Farrer 1903-06, 1:225-226.

  26. The information about Fordyce is taken from the Dictionary of National Biography, 19:432-433.

  end p.119

  consequence in that instance & he guessed at it as near as he could" (Chaldecott 1979, 74). Air thermometers did not have the same problem, but they generally required other problematic materials, such as glass, and mercury itself (for the measurement of pressure), which would soften, melt, boil, and evaporate.

  What was most commonly known as pyrometry at the time was based on the thermal expansion of various metals,27 but all of those metals melted in the higher degrees of heat found in pottery kilns and, of course, in foundries. Newton ([1701] 1935, 127) had estimated some temperatures higher than the boiling of mercury, but that was done by his unverified "law of cooling," which allowed him to deduce the initial temperatures from the amount of time taken by hot objects to cool down to certain measurable lower temperatures. Newton's law of cooling could not be tested directly in the pyrometric range, since that would have required an independent pyrometer. This was not a problem unique to Newton's method, but one that would come back to haunt every pyrometric scheme. It is very much like the problem of nomic measurement discussed in "The Problem of Nomic Measurement" in chapter 2, but even more difficult because the basis of sensation is almost entirely lacking in the extreme temperatures. (These issues will be considered further in the analysis part of this chapter.)

  Initially, in 1780, Wedgwood experimented with certain clay-and-iron-oxide compositions that changed colors according to temperature. However, soon he found a more accurate and extensive measure in a different effect, which is called "burning-shrinkage" in modern terminology: In considering this subject attentively, another property of argillaceous [claylike] bodies occurred to me; a property which … may be deemed a distinguishing character of this order of earths: I mean, the diminution of their bulk by fire. … I have found, that this diminution begins to take place in a low red-heat; and that it proceeds regularly, as the heat increases, till the clay becomes vitrified [takes a glassy form]. (Wedgwood 1782, 308-309)

  These pyrometric pieces were extremely robust, and their shrinkage behavior answered the purpose even better than Wedgwood had initially hoped.28 Their final sizes were apparently only a function of the highest temperatures they had endured, depending neither on the length of time for which they were exposed to the heat,29 nor on the lower temperatures experienced before or after the peak; "in three minutes or less, they are perfectly penetrated by the heat which acts upon them, so

  27. See, for example, Mortimer [1735] 1746-47, Fitzgerald 1760, and De Luc 1779.

  28. Years later Guyton de Morveau, whose work I will be examining in detail later, expressed similar surprise at how well the clay pieces worked. Guyton (1798, 500) tried two pyrometer pieces in the same intense heat for half an hour, and was impressed to see that their readings differed very little (160° and 163.5° on Wedgwood's scale): "I confess I did not expect to be so completely successful in this verification." See the end of "Ganging Up on Wedgwood," for Guyton's further defense of the comparability of the Wedgwood pyrometer.

  29. Chaldecott (1979, 79) notes that we now know this claim to be false, but that the proof of its falsity was not forthcoming until 1903.

  end p.120

  as to receive the full contraction which that degree of heat is capable of producing." So the pieces could be left in the hot places for any reasonable amount of time, and taken out, cooled, and measured at leisure afterwards (Wedgwood 1782, 316-317). This was an entirely novel method of pyrometry, and at that time the only one usable at extremely high temperatures. It was an achievement that propelled Wedgwood, already the master "artist," also into the top ranks of the (natural) "philosophers." Not only was Wedgwood's first article on the contraction pyrometer published in the Philosophical Transactions of the Royal Society in 1782, communicated by no less than Joseph Banks, the president, but it was also well enough received to facilitate Wedgwood's prompt election as a Fellow of the Royal Society.

  In that article, Wedgwood made no secret of his initial motivations: In a long course of experiments, for the improvement of the manufacture I am engaged in, some of my greatest difficulties and perplexities have arisen from not being able to ascertain the heat to which the experiment-pieces had been exposed. A red, bright red, and white heat, are indeterminate expressions … of too great latitude. …

  In the absence of a thermometer, he had relied on very particular benchmarks: Thus the kiln in which our glazed ware is fired furnishes three measures, the bottom being of one heat, the middle of a greater, and the top still greater: the kiln in which the biscuit ware is fired furnishes three or four others, of higher degrees of heat; and by these I have marked my registered experiments. (Wedgwood 1782, 306-307)

  But these measures were inadequate, and clearly unusable to anyone not baking clay in Etruria. In contrast, using the contraction of standard clay pieces had sufficient promise of quantifiability and wider applicability.

  Wedgwood (1782, 309-314) gave detailed instructions on the preparation of the clay, to be formed in small rectangular shapes (0.6 inch in breadth, 0.4 inch deep, and 1 inch long), and the brass gauge for measuring the sizes of the shrunken pieces. He attached a numerical scale assigning 1 degree of heat to contraction by 1/600 of the width of a clay piece. He acknowledged that "the divisions of this scale, like those of the common thermometers, are unavoidably arbitrary." However, Wedgwood was confident that the procedures specified by him would ensure comparability (as defined in "Regnault: Austerity and Comparability" in chapter 2): By this simple method we may be assured, that thermometers on this principle, though made by different persons, and in different countries, will all be equally affected by equal degrees of heat, and all speak the same language: the utility of this last circumstance is now too well known to need being insisted on. (314-315)

  With this instrument, Wedgwood succeeded in attaching numbers to a vast range of high temperatures where no one had confidently gone befor
e (318-319). The pyrometer quickly taught him many interesting things because it enabled clear comparisons of various degrees of heat. Wedgwood found that brass melted at 21 degrees on his scale (which I will write as 21°W), while the workmen in brass

  end p.121

  Figure 3.3. The pyrometer that Wedgwood presented to King George III (inventory no. 1927-1872). Science Museum/Science & Society Picture Library.

  foundries were in the habit of carrying their fires to 140°W and above; clearly this was a waste of fuel, for which he saw no purpose.30 He also found that an entire range from 27°W upwards had all been lumped together in the designation "white heat," the vagueness of which was obvious considering that the highest temperature he could produce was 160°W. Regarding his own business, Wedgwood (1784, 366) confessed: "Nor had I any idea, before the discovery of this thermometer, of the extreme difficulty, not to say impracticability, of obtaining, in common fires, or in common furnaces, an uniform heat through the extent even of a few inches."

  Wedgwood's pyrometer was an instant success, popular among technologists and scholars alike. For example, the Scottish chemist John Murray wrote in his textbook of chemistry (1819, 1:226): "The pyrometer which has come into most

  30. Wedgwood's enlightenment on this point may have been a shallow one, however. As noted by Daniell (1830, 281): "When metals are melted for the purposes of the arts, they of course require to be heated very far beyond their fusing points, that they may flow into the minutest fissures of the moulds in which they are cast, notwithstanding the cooling influences to which they are suddenly exposed. In some of the finer castings of brass, the perfection of the work depends upon the intensity to which the metal is heated, which in some cases is urged even beyond the melting point of iron."

  end p.122

  general use is that invented by Mr Wedgwood." Although Guyton de Morveau (1811b, 89) was critical of Wedgwood's instrument as we shall see later, he was clear about its proven utility and further potential: "I stressed the benefits that one could draw from this instrument, as already testified by the routine use of it by most of the physicists and chemists who make experiments at high temperatures." Lavoisier's collaborator Armand Séguin wrote to Wedgwood about the "greatest use" and "indispensability" of the Wedgwood pyrometer for their experiments on heat. Numerous other positive appraisals can be found very easily. There is an impressive list of chemists and physicists who are on record as having received pyrometer sets from Wedgwood, including Black, Hope, Priestley, Wollaston, Bergman, Crell, Guyton de Morveau, Lavoisier, Pictet, and Rumford. Wedgwood was proud enough to present a pyrometer to George III, which is the very instrument now preserved in the Science Museum in London (figure 3.3).31

  It Is Temperature, but Not As We Know It?

  Toward the end of his article Wedgwood indicated the next step clearly (1782, 318): "It now only remains, that the language of this new thermometer be understood, and that it may be known what the heats meant by its degrees really are." This desire was certainly echoed by others who admired his invention. William Playfair32 wrote Wedgwood on 12 September 1782: I have never conversed with anybody on the subject who did not admire your thermometer … but I have joined with severall [sic] in wishing that the scale of your thermometer were compared with that of Fahrenheit … [so] that without learning a new signification [or] affixing a new idea to the term degree of heat we might avail ourselves of your useful invention. (Playfair, quoted in McKendrick 1973, 308-309)

  Wedgwood did his best to meet this demand, apparently helped by a useful suggestion from Playfair himself. Wedgwood opened his next communication to the Royal Society as follows: This thermometer … has now been found, from extensive experience, both in my manufactories and experimental enquiries, to answer the expectations I had conceived of it as a measure of all degrees of common fire above ignition: but at present it stands in a detached state, not connected with any other, as it does not begin to take place till the heat is too great to be measured or supported by mercurial ones. (Wedgwood 1784, 358)

  To connect up his pyrometric scale and the mercury-based Fahrenheit scale, Wedgwood looked for a mediating temperature scale that would overlap with both. For that purpose he followed the familiar pyrometric tradition of utilizing the expansion of metals, and he chose silver. Wedgwood's patching-up strategy is

  31. For Séguin's letter, the list of those who received pyrometer sets from Wedgwood, and other records of the esteem that Wedgwood's pyrometer enjoyed among his contemporaries, see Chaldecott 1979, 82-83, and McKendrick 1973, 308-309.

  32. This was probably William Playfair (1759-1823), Scottish publicist who later participated in the French Revolution, and brother of John (1748-1819), the geologist and mathematician.

  end p.123

  Figure 3.4. Wedgwood's patching of three temperature scales; the upper portion of the Fahrenheit measure is not known to begin with. Source: Wedgwood 1784, figs. 1, 2, and 3, from plate 14. Courtesy of the Royal Society.

  end p.124

  represented graphically in figure 3.4, which he originally presented to the Royal Society "in the form of a very long roll."

  The basis of this Wedgwood-Fahrenheit comparison was quite straightforward (Wedgwood 1784, 364-369). The expansion of the silver piece was measured in a gauge very much like the one he used to measure the clay pyrometer pieces. The low end of the silver scale overlapped considerably with the mercury scale. He set the zero of the silver scale at 50°F and found that the expansion of silver from that point to the heat of boiling water (212°F) amounted to 8 (arbitrary) units on the gauge; so an interval of 8° on the silver scale corresponded to an interval of 162° on the mercury-Fahrenheit scale. Assuming linearity, Wedgwood concluded that each silver degree "contained" 20.25 mercury-Fahrenheit degrees, simply by dividing 162 by 8. Putting this result together with a similar one obtained between 50°F and the boiling point of mercury, Wedgwood arrived at an approximate conversion ratio of 20 mercury-Fahrenheit degrees for each silver degree. Similar calculations at the high end of the silver scale gave the conversion ratio between the Wedgwood scale and the silver scale as 6.5 silver degrees for each Wedgwood degree.33 Putting the two conversion factors together, Wedgwood found that each Wedgwood degree was worth 20 × 6.5 = 130 Fahrenheit degrees. From that conversion ratio and the information that the zero of the silver scale was set at 50°F, he also located the start of his scale (red heat) at 1077.5°F.34 Now Wedgwood had achieved his stated aim of having "the whole range of the degrees of heat brought into one uniform series, expressed in one language, and comparable in every part" (1784, 358). He presented a double series of temperatures in Fahrenheit and Wedgwood degrees, some of which are shown here in table 3.1.

  Wedgwood's measurements were the first concrete temperature values with any sort of credibility in the range above the upper limit of the mercury thermometer. However, much as his work was admired, it also drew increasingly sharp criticism. First of all, there were enormous difficulties in reproducing his clay pyrometer pieces exactly. His initial desire was that others would be able to make their own Wedgwood thermometers following his published instructions, but that did not turn out to be the case.35 As Wedgwood himself acknowledged (1786, 390-401), the properties of the clay pieces depended on intricate details of the process by which they were made, and producing uniformly behaving pieces turned out to be very difficult. The making of the standard clay pieces required "those peculiar niceties and precautions in the manual operations, which theory will not suggest, and which practice in the working of clay can alone teach" (quoted in Chaldecott 1979, 82). There was also the need to use exactly the same kind of clay as used for the original pieces. Initially Wedgwood was sanguine about this problem (1782, 309-311), thinking that

  33. In a furnace, when the Wedgwood pyrometer indicated 2.25°, the silver thermometer gave 66°; in another instance, 6.25° Wedgwood corresponded to 92° silver. Hence an interval of 4° on Wedgwood's thermometer was equivalent to an interval of 26° on the silver thermomete
r, which gives 6.5 silver degrees for 1 Wedgwood degree.

  34. Wedgwood made the calculation as follows: 2.25°W = 66° silver = (50°F +66 × 20°F) = 1,370°F; so 0°W = (1,370°F − 2.25 × 130°F) = 1077.5°F.

  35. This was noted particularly in France; see Chaldecott 1975, 11-12.

  end p.125

  Table 3.1. Some high temperatures measured by Wedgwood, with conversion to Fahrenheit degrees Phenomenon

  °Wedgwood

  °Fahrenheit

  Greatest heat of Wedgwood's air-furnace

  160

  21877

  Cast iron melts

  130