Evolution of the Thermometer/Chapter 3
III. Attempts to Obtain a Standard Scale from Boyle to Newton.
Through whom knowledge of the thermometers devised by the Florentine Academy reached England is not known, but it has been suggested that the French traveler Monconys conveyed it to the Hon. Robert Boyle on the occasion of his visit to London in 1663, and there is circumstantial evidence in favor of this view. Monconys was most politely received by the scholarly Irishman and attended a meeting of the Royal Society on the 30th of May; he had with him in London one of the new instruments and made an entry in his diary on the 31st May to this effect: "The weather was cold towards evening and the thermometer fell to 6.5 degrees."
While the Accademia del Cimento was busy experimenting on heat and cold, magnetism and acoustics, and trying to prove the non-existence of positive levity," the British philosopher was working in similar fields; he improved the air-pump (invented in 1650 by Otto de Guericke), devised a physico-mechanical experiments touching the spring of the air," discovered the fundamental truth known as "Boyle's Law," invented the manometer, and made a great variety of observations in chemistry and physics of prime importance. All this work qualified him for thermometrical studies, and it is said that he constructed a "sealed weather-glass" before he saw the Italian instrument, but this is improbable.
Boyle graduated the stems of thermometers with "little specks of amel" into inches and fractions as small as sixteenths; in one experiment he found that "sal-armoniac" dissolved in water "made it descend to 2-11/16 inches in a quarter of an hour. He observed that thermometer stems were not sufficiently even and cylindrical, being often widest near the bulb, and said this was a source of inaccuracy.
Boyle felt the need of a standard permitting comparison of effects shown by different thermometers, and expressed it thus: "We are greatly at a loss for a standard whereby to measure cold. The common instruments show us no more than the relative coldness of the air, but leave us in the dark as to the positive degree thereof; whence we cannot communicate the idea of any such degree to another person. For not only the several differences of this quality have no names assigned them, but our sense of feeling cannot therein be depended upon; and thermometers are such very variable things that it seems morally impossible from them to settle such a measure of coldness as we have of time, distance, weight, etc." (1665).
Boyle endeavored to overcome this difficulty; believing that the melting-point of ice varied with geographical latitude, he proposed using the oil of aniseed for getting a fixed point, placing it around the bulb of an alcohol thermometer, allowing the oil to freeze and marking the height of the spirit of wine in the bulb "when the oil begins to curdle."
This scheme for getting a fixed point has been wholly misunderstood by some historians who state that Boyle filled his thermometers with aniseed oil!
While Boyle tried to secure one fixed point he overlooked the advantages of having two; he strove to compute the absolute expansion of alcohol and to divide the scale into ten thousandths, or some aliquot part of the total expansion.
Strange notions of natural phenomena were current in Boyle's day and the "Father of Chemistry " was not above crediting absurdities; he quoted Orthelius who wrote: "The liquor distilled from the ore of magnesia, or of bismuth, will swell considerably in the glass it is kept in at the full moon, and subside at the new."
Contemporary with Boyle, another distinguished British philosopher was occupied with improvements in thermometers. Robert Hooke, afterwards secretary of the Royal Society, published in 1664 his "Micrographia;" in this work he says: "I have brought sealed thermometers to a great certainty and tenderness, for I have made some with stems above four feet long in which the expanding liquor would so far vary as to be very neer the top in the heat of summer and preety neer the bottom at the coldest time of winter." Hooke filled his thermometers with "best rectified spirit of wine highly ting'd with the lovely colour of cochineal." To graduate the stem he placed zero at the point which the liquid stood when the bulb was placed in freezing distilled water; he then marked the divisions above and below "according to the degrees of expansion or contraction of the liquor in proportion to the bulk it had when it indur'd the freezing cold."
Edmund Halley, writing in 1700, said that Hooke exhibited at Gresham College, 2nd January, 1667-8, a combination of barometer and thermometer in separate tubes, the freezing-point of water being equal to zero, and the stem being graduated from -70 to +130.
Hooke made an important step in advance when he took the freezing-point of water as a fixed point in the scale, but the statement made by Brewster that in 1684 Hooke proposed the boiling-point as a second fixed point has been examined by Poggendorff and not verified. The claim has been made that the two fiduciary points were first proposed by the Dutch mathematician Christian Huyghens, in a letter dated 2nd January, 1665, addressed to Robert Moray. (A. Momber, Schr. naturf. Gesch. Danzig, N. F. VII, 108.) Huyghens wrote: "It would be well to have a universal and determinate standard for heat and cold, securing a definite proportion between the capacity of the bulb and the tube, and then taking for the commencement the degree of cold at which water begins to freeze, or better the temperature of boiling water, so that without sending a thermometer to a distance, one could communicate the degrees of heat or of cold found in experiments and record them for the use of posterity." In this passage Huyghens does indeed suggest the two phenomena for fixing a standard, but only as alternatives, and he seems to have had no idea of dividing the space between them.
The proposition to divide into equal parts the interval between two points to be ascertained by experiment was made four years later by Honoré Fabri, a Jesuit of French birth, who had been one of the corresponding members of the Accademia del Cimento. In his voluminous work on physics published in 1669, he describes an experiment with a Florentine thermometer for the purpose of constructing such a scale; he applied snow in very cold weather to the bulb and marked the point at which the liquid stood, then he marked the position of the liquid at the highest heat of summer and divided the line drawn between these points into eight equal parts. As we now know, the higher fixed point was ill-chosen, but the method was correct in principle, though not adopted until long after. Meteorological observations were made with Hooke's thermometers by John Wallis, professor of mathematics in Oxford University, and he recorded in a certain paper published in the Philosophical Transactions for 1669, (p. 113), that the "liquor" stood at three and one-half inches on December 26, 1669, and at seven inches in "brisk frosts."
Several novel forms of thermometers were constructed about 1660-62 by the accomplished experimenter in physics, Otto de Guericke, Burgomaster of Magdeburg; they all bore impress of his genius, and one of them was, and still remains, unique in many particulars. It was gigantic in size being above twenty feet long, gorgeous with blue paint and gilt stars, and decorated with the image of a winged angel whose outstretched arm pointed to the temperature at every moment. For convenience the immense air-thermoscope was fastened to the wall of a house on the shady side; it was renowned for its power of showing "the coldest and hottest weather throughout an entire year." The instrument consisted of a large copper globe joined to a long tube one inch wide, of the same metal; the tube was bent upon itself
so as to form a very narrow U, in which was placed a certain amount of alcohol. The shorter arm of the U was open at the top; on the liquid within it floated a tiny inverted cup of brass foil to which a cord was attached that passed around a wheel, hung upon the underside of the globe, and carried at the other end a little figure of an angel pointing to the scale on the tube; the tube was concealed from the observer by a wooden case, and the image hung without. A valve at one side of the large copper sphere permitted enough air to be withdrawn by means of the air-pump, to adjust the height of the image, which hung about half way up the tube. On the fifteen-foot scale were the words: "Magum frigus, aer frigidus, aer subfrigidus, aer temperatus, aer subcalidus, aer calidus, magnus calor;" and the large sphere above the tube was inscribed "Mobile perpetuum."
De Guericke constructed a barometer on similar lines, and gave it the legend "Semper vivum;" he described it in a letter to G. Schott, dated 30 December, 1661.
Still more remarkable was the self-registering meteorological apparatus devised by the great German physicist; it recorded every hour the change in temperature, direction of the wind, the rainfall, and amount of snow or hail. Monconys describes this briefly, with a diagram showing the arrangement, in his Journal des Voyages (Vol. III, 1663).
The differential thermometer was invented by Gaspar Schott, as early as 1657 (Mechan. hydraul-pneumat. II, 231), and afterwards improved by Joh. Christ. Sturm, Professor in Altorf, Bavaria, in 1676. It was a U-tube with arms of uneven length, both tubes being closed; Sturm explained its action quite correctly.
In the latter part of the seventeenth century references in scientific literature to the construction and use of the ordinary thermoscopes multiplied; Le Febure in his admirable treatise on chemistry (1669), mentions the thermometer as a well-known instrument, and Gabriel Clauder in the "Miscellanea curiosa Acad. nat. curiosa" (Dec. 2, Anno 6, p. 351, 1687), describes a "thermoscopium noviter inventum " adapted to immersion in liquids.
There was published at Amsterdam, in 1688, an illustrated work entirely devoted to barometers, thermometers, and hygrometers, written by Dalencé, who concealed his name under the initial D——. I have already referred to this interesting book in connection with the Drebbel myth, but it deserves emphasizing, for it contains an imperfect summary of thermometrical knowledge up to that date; to avoid repetition, however, I shall only notice items not previously given here.
Dalencé describes the Italian thermoscope as having a bulb the size of a pigeon egg and a tube as big as a quill pen; he suggests the use of a mixture of three parts of water with one of aqua fortis to prevent the liquid freezing, and a flattened bulb to permit heat, or cold, more readily to penetrate the centre of the liquid. He proposed, also, that two points should be marked on the scale, the freezing-point of water, to be marked "cold," and the melting-point of butter; the space between them to be divided equally and the centre to be marked "temperate." Then each space above and below "temperate" to be divided into ten equal degrees, four additional degrees to be placed above the melting-point of butter and four below the freezing-point of water, making a scale of thirty degrees in all.
|Melting butter …||10||86°|
|Freezing temperature …||−10||32°|
He further proposed another standard scale, the fixed points to be the temperature of a cellar and of ice, the space between to be divided into fifteen degrees; and he added: "all thermometers made by the latter method are comparable." To make instruments easy of transportation, and for fancy, the stems were bent into circular, oval, spiral, triangular, and stellate shapes, and Dalencé gives figures of each. He also described the floating glass bulbs of. Kircher, and made them in the shape of turtles to apply to the arms and body of feverish persons. "Some curieux" he says, "use mercury in thermometers," but the instrument he writes of was an air-thermoscope and the fluid metal was not employed on account of its property of expansion. Dalencé praised the skill of Sieur Hubin, glass-blower, whose address was Rue St. Martin, Paris, and says his success is due to the fact that "he knows the reasons for that which he does."
The sections in Dalencé's work on the barometer and hygrometer are interesting from the historical point of view, but do not fall within the province of these chapters.
Dalencé seldom gives credit to individuals for their shares in the development of the instruments he describes, and it is difficult to determine how many of the improvements mentioned by him were original with him, probably but few.
Mention may be here made of a complicated thermometer constructed by the expert Parisian glass-blower Hubin, although it was not described in print until 1725 (Reyher, Pneumatica). Two bulbed tubes were united at a reservoir, and bent in the shape of an U; both tubes were closed so that the apparatus was independent of air-pressure. The shorter arm of the U was filled with mercury up to the centre of the reservoir, the longer arm half filled with water, the remaining half, including the large bulb, containing air. When the air in the bulb expanded, it pressed upon the column of water which in turn forced the mercury up the shorter and wider tube. Hubin claimed for this instrument greater delicacy than the Florentine in the proportion of 216 to 4, as determined by experiment.
The question, who first used quicksilver as the dilating liquid in thermometers, is apparently a simple one, to be easily answered, but like many other questions of priority in the history of the thermometer, many claims have been advanced and the problem requires examination; no less than ten names are mentioned by different authorities as inventors of mercury thermometers. Thermometers containing mercury were indeed made at an early date, but the liquid metal was only an accessory to the air-thermoscope and was not used as a heat-measurer.
Athanasius Kircher, in 1643, mentions a thermoscope containing mercury, but does not describe the function of the liquid metal. The Accademia del Cimento, as related in the diary of the society, made mercury thermometers in 1657, compared them with water-thermometers of the same size, and observed that the former fell and rose more quickly than the latter when affected by changes of temperature, although the total amount was less; but this important fact lay dormant and unused.
Mercury thermometers were also known in Paris; Ismael Boulliau is said to have made them in 1659; and a letter dated 28th May, 1684, written in Paris to the Royal Society, London, by Mr. Musgrave, describes one three inches long and five lines in diameter, that was used for taking the temperature of fever patients.
Christian Huyghens, in 1665, and Dalencé, in 1668, are also credited with the invention of mercury thermometers. In the same year that Dalencé's book appeared, Edmund Halley, the eminent English mathematician and astronomer, was studying experimentally the relative expansion of water, alcohol and mercury, and he stated that mercury would make a good thermometrical liquid if its coefficient of expansion was greater; he did not actually recommend the liquid metal, although he perceived that its expansion was large enough to influence the readings of the barometer. He observed that mercury heated in boiling water ceased to expand on long continuing the operation, and according to Momber he proposed taking the boiling-point of water as a fixed point, but Poggendorff says he made no application of this fact to thermometry. Halley thought that "spirit of wine" lost part of its expansive force by long keeping, yet he proposed the boiling-point of alcohol as a limit to thermometrical scales. The English scientist felt the need of a reliable thermometer scale and expressed himself in somewhat the same way as his countryman Boyle had done, fourteen years before: "I cannot learn that any thermometer of either sort was ever made or adjusted so as it might be concluded what the degrees or divisions of the said instrument did mean; neither were any thermometers ever otherwise graduated but by standards kept by each particular workman without any agreement or reference to one another. So that whenever observations of a thermometer are made by any curious person to signify the degree of heat in the air or other thing they cannot be understood, unless by those who have by them thermometers of the same make and adjustment."
Halley regarded the freezing-points of aniseed oil, and of water, as unreliable, and preferred as a starting point the temperature of a deep cellar the constancy of which had been demonstrated by De la Hire in the crypt of the Paris observatory.
Christian Wolf, of Halle, and Olof Römer, of Copenhagen, both in 1709, are also named as the first to use mercury as a heat measuring liquid, but in spite of these many claimants the fact remains that Fahrenheit, in 1714, was incontestably the first to construct mercury thermometers having reliable scales. But this anticipates.
The need of a standard scale, easily made and based on constant phenomena that can be reproduced at will, was felt by all who used thermometers, and an important practical proposal to secure this desideratum was made in 1694 by Carlo Renaldini, a former member of the Accademia del Cimento. and professor of mathematics in Padua. At that date, and in the eightieth year of his age, he published a work on natural philosophy, in which he suggested taking the melting-point of ice and the boiling-point of water for two fixed points of thermometer scales, and dividing the space between them into twelve equal parts. This truly admirable proposition was not appreciated by his contemporaries who did not wholly believe in the constancy of these temperatures, and it was forgotten by succeeding philosophers, thus delaying greatly accurate observations of temperature.
Dalencé had anticipated Renaldini in adopting the principle of the subdivision of the interval between two determinable points, but the phenomena chosen by the Frenchman were not so reliable as those proposed by the Italian, which were afterwards adopted by Celsius.
Renaldini devised another method for graduating thermometers; he plunged the thermometer to be graduated first in mashed ice, then in a mixture of eleven parts cold water plus one of boiling water, and successively in mixtures of ten cold plus two boiling, nine cold plus three boiling, eight cold plus four boiling, and lastly in boiling water itself, marking on the scale the position of the expanding fluid at each immersion. This sounds plausible, but was shown by Wolff to be deceptive and unreliable.
Sir Isaac Newton was one of those who attacked the thermometrical problem of the age, but the scale proposed by this great genius was by no means satisfactory; he rejected alcohol as the dilating liquid and preferred "lintseed" oil; for fixed points in the scale he chose the temperature of melting snow and of the human body, dividing the interval into twelve equal parts. In his paper "Scala graduum," published anonymously in the Philosophical Transactions, May, 1701, Newton gave his method of graduation; he assumed that when the instrument was placed in melting snow the linseed oil occupied 10,000 parts, and found that the same oil at the temperature of the human body, which he called one degree of heat, occupied a space of 10,256 parts; in water boiling violently 10,725 parts, and in melted tin beginning to cool 11,516 parts; from this he computed the degrees of heat corresponding to the phenomena, calling the heat of the human body 12, he found for boiling water 34, and melting tin 72. His thermometer was three feet long and had a bulb two inches in diameter.
Newton also made an experiment with a thick piece of iron as a pyrometer; he heated it red hot and "put it in a cold place where the wind blew uniformly," then he placed on the bar particles of various metals and other fusible bodies and noted the times of cooling, until all the particles having lost their fluidity grew cold, and the heat of the iron was equal to that of the human body. Then by ming that the excesses of the heats of the iron and of the solidified substances, above the heat of the atmosphere, were in geometrical progression when the times were in arithmetical progression, all the heats were obtained. From this experiment and computation Newton drew up the following scale of degrees of heat.
|0||0||heat of the winter air when water begins to freeze.|
|1||12||greatest heat of the surface of the human body.|
|2||24||heat of melting wax.|
|2.5||34||heat of water boiling vehemently.|
|3||48||lowest heat at which equal parts of tin and bismuth melt.|
|4||96||lowest heat at which lead melts.|
|5||192||heat of a small coal fire not urged by bellows.|
Biot, commenting on Newton's paper, notes that it contains three important discoveries: (1) a method of making thermometers comparable by determining the extreme terms of their graduation from the phenomena of constant temperature; (2) the determination of the law of cooling in solid bodies at moderate temperatures; (3) observation of the constancy of temperature in fusion and ebullition.
From Newton's note-books it appears that he was occupied with these studies in March, 1692-3, although they were not made public until 1701.
Newton made no use of the boiling-point of water in constructing his scale of temperatures, as he considered it variable; he recorded that water begins to boil at 33° of his scale, and boils vehemently at 34° to 34½°. We now know that such fluctuations depend upon the position of the thermometer (which must not be immersed in the liquid), on the pressure of the atmosphere, on the chemical purity of the water, and on the shape of the vessel holding it, so it is not surprising that doubts existed as to the constancy of the phenomenon. Mariotte had laid the foundations of hypsometry, but the experimental proofs were not secured until Le Monnier tested the matter in the Pyrenees in 1739.
In the same year that Newton published his researches, Étienne François Geoffroy described an open air-thermoscope nearly identical with that of Athanasius Kircher, made in the year 1643, and to which the French savant made no allusion. (Phil. Trans., 1701, p. 951.)