The thermometer was naturally of interest to the Fellows of the new Royal Society, and there are references to it at about the time that Robert Boyle was writing his New Experiments and Observations Touching Cold. From one of the earliest of their records we learn that their ideas were not very highly developed. On October 7, 1663, they were discussing the provision of four thermometers, to be kept in cellars, and we read that, "For the adjusting of these thermometers it was thought sufficient to know what mark they stood at then; and for the making of observations, to take notice, how they should afterwards differ from this mark." Five days later Hooke was instructed to "make ready the two thermometers of Dr. [Christopher Wren's invention, one of tin, the other of glass," and on October 14 there is another mention of "Dr. Wren's new kind of thermometer with two round glasses and quicksilver in them.""3
One of these thermometers is clearly shown in a drawing bound in a copy of the Parentalia belonging to the Royal Institute of British Architects. It forms part of a meteorograph or "weather clock," certainly the first such instrument which recorded wind direction and temperature, as well as attempting to indicate hourly amounts of rainfall. The thermometer consisted of a large bulb connected by a narrow tube to a vertical cylinder partly full of mercury. A float in the latter is connected to a cord passing over a pulley, then horizontally over another pulley to a counterweight. A pencil is attached to the horizontal portion of the cord and records on a rectangular chart moved sideways by a clock. A close examination of Wren's drawing would suggest that this is an air thermometer, and indeed if the drawing is even approximately to scale, this must be so.
This meteorograph was seen by Balthasar de Monconys on June 10, 1663. Wren's second thermometer "with two round glasses and quicksilver in them" was also intended for a meteorograph." The only clue I have found to its construction is contained in a famous letter written on July 30, 1663, to Lord Brouncker, the President of the Royal Society,& suggesting experiments for the entertainment of Charles I on the occasion of a proposed visit that the King never made. The relevant passage is as follows:
I have pleased myself not a little with the play of the weather wheele (the Onely true way to measure Expansions of the Ayre) & I fancy it must needs give others satisfaction if it were once firmely made which I suppose may be donne if the circular pipe which cannot be truly blown in glasse were made of Brasse by those who make Trumpets & Sackbuts (who wire draw their pipes through a hole & equal them, & then filling them with melted Lead turne them round into what flexures they please) but the inside of the pipe must be vernished with China vernish (which Gratorex hath) to preserve it from the quicksilver, & the glasses must be fixed to the Pipe with Varnishe, which I take to be the best cement in the world, for thus the Chinese fix glasse & mother of Pearle in their workes.
If we consider the instrument illustrated in Fig. 2.6, this passage leaves us with a poor impression of the state of glassworking in England in the seventeenth century.
Hooke and Wren were very good friends and exchanged ideas freely, 1 and Hooke was also interested in constructing a meteorograph, finishing an elaborate one about fifteen years later.11 He had a scheme for a thermograph on January 4, 1664/5, for we read that on that day
Mr. Hooke shewed the way of applying a thermometer to a weathercock [sic], by sealing up spirit of wine in a glass cane, with two pretty large heads, one of which was filled with spirit of wine, as was also the intermediate stem; the other not quite full, a space of air being left to give liberty for the expanding liquor. The cane thus filled was poised in the manner of the beam of a balance, and fits operation was described]. It was moved to consider, whether this instrument were sensible and nice enough. 12
The same doubts seem to have arisen about another somewhat different instrument, for on the twenty-seventh of the same month we find Sir Robert Moray writing to Christian Huygens at The Hague: "Nous avons des Thermometres faits en spiral qui tournent autour dun Essieu, dont le moindre Movement marque sur un grand cercle un grande espace. Mais il n'est pas assez delicat."1 The mention of a "spiral" reminds one of Drebbel's "perpetual motion" as described by Antonini,1 but if it had been an air thermometer the question of sensitivity would have been less likely to arise. It is of some interest that in 1878 and 1879 patents13 were awarded to Friedrich Kuntze of Leipzig for a "Drehthermometer" of this general sort, but more highly developed technically.
These mechanical schemes, however interesting, are not our main concern. The Royal Society soon recognized the importance of making comparable thermometers; and when, on October 21, 1663, Hooke brought out a number of spirit thermometers and said they all agreed "at the figure 8," it was decided "that one of them should be kept by the Society as a standard, and all the rest so adjusted, as to agree therewith."1 And on December 2 of the same year, Dr. Jonathan Goddard was asked to "draw the air out of" rectified spirit of wine and see how it worked in a thermometer.17 We shall see that for another century people tried this without realizing that they were drawing a good deal of the alcohol out at the same time.
It must have been at about this period that Robert Hooke devised the earliest method of preparing comparable thermometers from first principles, without the necessity of making their dimensions exactly similar as the Florentines had tried to do. Indeed, the thermometers just referred to may well have been made by Hooke's new method, which was first published two years later in his remarkable Micrographia, 1s a book chiefly devoted to the microscope and the things that can be seen with its aid.
Hooke first describes the construction of a large thermometer -as much as four feet long- and its filling with "the best rectified spirit of wine highly ting'd with the lovely colour of cocheneel." He chose spirit of wine for three reasons: because it is easily colored; because it responds quickly to heat; and because "it is not subject to be frozen by any cold yet known." He then deals with the graduation of his thermometer, and we shall quote his own words:
Then, for graduating the stem, I fix that for the beginning of my division where the surface of the liquor in the stem remains when the ball is placed in common distilled water, that is so cold that it just begins to freeze and shoot into flakes; and that mark I fix at a convenient place of the stem, to make it capable of exhibiting very many degrees of cold, below that which is requisite to freeze water: the rest of my divisions, both above and below this (which I mark with a [0]1 or nought) I place according to the degrees of expansion, or contraction of the liquor in proportion to the bulk it had when it indur'd the newly mention'd freezing cold. And this may be very easily and accurately enough done by this following way; prepare a cylindrical vessel of very thin plate brass or silver, ABCD of [Fig. 3.11: the diameter AB of whose cavity let be about two inches, and the depth BC the same; let each end be cover'd with a flat and smooth plate of the same substance, closely soderd on, and in the midst of the upper cover make a pretty large hole EF, about the bigness of a fifth part of the diameter of the other; into this fasten very well with cement a straight and even cylindrical pipe of glass, EFGH, the diameter of whose cavity let be exactly one tenth of the diameter of the greater cylinder. Let this pipe be mark'd at GH with a diamant, so that G from E may be distant just two inches, or the same height with that of the cavity of the greater cylinder, then divide the length EG exactly into 10 parts, so the capacity of the hollow of each of these divisions will be ¼00 part of the capacity of the greater cylinder. This vessel being thus prepared, the way of marking and graduating the thermometers may be very easily thus performed: Fill this cylindrical vessel with the same liquor wherewith the thermometers are fill'd, then place both it and the thermometer you are to graduate, in water that is ready to be frozen, and bring the surface of the liquor in the thermometer to the first marke or [0]; then so proportion the liquor in the cylindrical vessel, that the sur face of it may just be at the lower end of the small glass-cylinder; then very gently and gradually warm the water in which both the thermometer and this cylindrical vessel stand, and as you perceive the ting'd liquor to rise in both stems, with the point of a diamond give several marks on the stem of the thermometer at those places, which by comparing the expansion in both stems, are found to correspond to the divisions of the cylindrical vessel, and having by this means marked some few of these divisions on the stem, it will be very easie by these to mark all the rest of the stem, and accordingly to assign to every division a proper character. 20
This is the earliest serious attempt at a scale of temperature that could really be standardized. As usual, Hooke's efforts in this field were forgotten, and Réaumur could describe his rather similar method in the next century with no reference to them at all. While it was an excellent attempt, it suffered from several disabilities: spirit of wine is not a well-defined substance, its properties varying rapidly with the amount of water in it; Hooke's choice of the freezing point of water, rather than the melting point of ice, was unfortunate; and he did not make any allowance for the difference in expansion of brass--or silverand glass. It may be noted that Réaumur inherited the first two of these sources of error.
At any rate, here we have an example of one of the two ways of establishing the scale of a liquid-in-glass thermometer. The other, using two fixed points, was finally recognized as being far superior.
Hooke's thermometer is the subject of an interesting study by Louise Diehl Patterson, who lists a number of celebrated users of this scale between 1668 and 1708-Boyle, Derham, Halley, Hauksbee, Hunt, and Locke. From the dimensions of Hooke's apparatus and the coefficient of expansion of alcohol solutions, she deduces that in 1664 a degree of Hooke's scale was about 1.1 or 1.2°C. But from a consideration of the scales in Hooke's "marine barometer," a combination of an air thermometer and a spirit thermometer, 23 she concludes that he must soon have doubled the value of his degrees to about 2.38°C.24 Papin, Hunt, Derham, Hauksbee, and Halley used a unit ⅒ as large as Hooke's revised degree.
These results of Patterson's have been used by Professor Gordon Manley in a valiant attempt to correlate several seventeenthcentury records of London temperature with modern measurements.
Robert Boyle seems to have doubted whether spirit of wine is the best thermometric liquid, wondering whether "well-refined quicksilver' might not be superior.2 In this he was to some extent supported by Edmond Halley, who measured the dilatation of mercury from room temperature to the boiling point of water, and thought that "Mercury may possibly serve as well for a thermometer as most other liquors. "2 But the small expansion of mercury was a difficulty, and in 1693 Halley wrote that it "might most properly be applied to the construction of thermometers were its expansion more considerable."28 But even though the expansion of mercury is small it is enough to cause an error in the readings of the barometer. Spirit of wine would be a good thermometric liguid, except that it is variable in quality, and he suggests that the boiling point might be as good a test as any of the strength of spirit or spirituous liquors. After some discussion Halley concludes that air is very much the best fluid for thermometers.
None of this seems very important; but we are immediately brought up short by the following passage:
Now the thermometers hitherto in use are of two sorts; the one shewing the differing temper of heat and cold by the expansion of spirit of wine, the other by the air; but I cannot learn that any of them of either sort were ever made or adjusted, so as it might be concluded, what the degrees or divisions of the said instruments did mean; neither were they ever otherwise graduated, but by standards kept by each particular workman, without any agreement or reference to one another: so that . . . observations . . . cannot be understood, unless by those who have with them thermometers of the same make and adjustment. Much less has the way been shewn how to make this instrument without a standard, or to make two of them to agree artificially without comparing them together.