Slike strani
[blocks in formation]

0°; hence T,-T=100, and To=50°, and we have F=100(1598 - 5.82 X 50)=130,700. Dividing this by 100,000,000 to reduce it to volts, the final conclusion is, that a thermo-electric couple of the kind described will give an electromotive force of about 0.00131 of a volt (in round aumbers). To find the "neutral point” of an iron-copper couple, we merely have to set the expression 1598 - 5.82T, equal to zero, and solve the equation for To. Proceeding in this manner, we find that the desired neutral temperature is 274° C.

[blocks in formation]
[ocr errors]

Nickel is not included in the foregoing table, because its behavior is anomalous. From - 18° C. to 175° C. its constants are B=-2204 and C=-5.12; but a short distance above 175° C. their values change profoundly, so that between 250° C. and 310° C. we have B = -8449 and C=+24.1. Above 340° C. we have, for this metal, the values B 307 and C= - 5.12. Antimony and bismuth are commonly used as the two metals in experimental thermo-electric circuits, since they yield an electromotive force which is larger than is obtainable under similar conditions by other metallic pairs. The thermoelectric constants of antimony and bismuth do not appear to have been determined with precision, however. When the average temperature of the two junctions (T.) is about 20° C., the values of B + CT, are approximately as follows: Bismuth, pressed commercial wire, - 9700; bismuth, pure pressed wire, -8900; bismuth crystal, axially, —-6500; bismuth crystal, equatorially, - 4500. Antimony, pressed commercial wire, + 600; antimony, pure pressed wire, + 280; antimony crystal, axially, + 2260; antimony crystal, equatorially, + 2640.

When a thermo-electric couple is generating an electrical current, heat is absorbed at the hot junction, and given out at the cold one.

If an electric current is caused to flow across the junction of any two dissimilar metals, heat is either evolved or absorbed at the junction; and if an evolution of heat is observed when the current flows across from the metal X to the metal Y, there will be an absorption of heat when the current is made to flow from Y to X. This phenomenon, which was discovered by Peltier, is known as the Peltier effect.” Lord Kelvin showed that in a thermo-electric circuit the reversible heat effects are not confined to the junctions themselves. He showed, in fact, that when an electrical current is caused to flow through a wire that is locally heated by a gas flame or otherwise, the current tends to

cause a displacement, along the length of the wire, of the point of maximum temperature. The effect may be illustrated by thinking of the wire as a tube conveying a stream of some real fluid; the fluid taking up more or less of the heat, and carrying it along in such a way as to shift the point at which the temperature is greatest. The analogy is imperfect, however, because in the case of electricity the heat travels with the current in some metals, and against it in others. This phenomenon (which is known as the «Thomson effect”) has an important bearing upon the theory of thermoelectricity, because in a thermo-electrical circuit the temperature is necessarily variable from point to point, and, therefore, the «Thomson effect) may (and in fact does) modify the phenomena considerably. The effect is zero in lead, but in nearly every conductor it is quite sensible. The consequences of the (Thomson effect” are discussed in two papers in the first volume of Lord Kelvin's Mathematical and Physical Papers.'

When a circuit contains several thermoelectric couples, with their successive junctions alternately heated and cooled, the total electromotive force that is produced is sensibly equal to the sum of all the electromotive forces that the several couples would produce, if each existed separately. Advantage is taken of this fact in the instrument known as the thermopile, which consists of a series of small bars of antimony and bismuth (often about 25 pairs altogether), disposed side by side so as to form an approximate cube and insulated from one another by strips of paper or other nonconductor. These bars are soldered together at the ends so that a current, in order to traverse the system, must pass back and forth through alternate bars of bismuth and of antimony; the passage of the current from bismuth into antimony being always effected (let us say) at the right hand end of the little pile of bars, while the passage from antimony into bismuth is always effected at the left hand end. The two free ends of the thermo-pile being connected by a wire, a current of electricity flows through the circuit so formed, when either of the surfaces containing the soldered junctions is warmed, while the opposite one is kept cool. This instrument is used to a considerable extent for the detection and even for the approximate estimation of radiant heat; the current that it produces and which is measured by a delicate galvanometer in the external part of the circuit being taken as the index of the amount of the radiation. (Consult Tyndall

, Contributions to Molecular Physics in the Domain of Radiant Heat'). The thermo-pile was invented by Nobili, but it was so greatly improved by Melloni that it is commonly credited to him.

Many attempts have been made to construct a thermo-electric combination that would yield a current of electricity intense enough to be of commercial utility; but while such attempts have been partially successful, there is no great reason to suppose that thermo-electric generators will ever be of much practical value. The difficulties are partly structural and partly theoretical. In order to realize any considerable electromotive force, the number of elements must be large; and the experience heretofore



has been that a complicated thermo-pile is not eter depends for its action upon the fact that particularly durable. Even if this objection mercury expands about seven times as much as can be overcome, there remains the serious glass, for a given rise in temperature; so that one that there are theoretical reasons, based when mercury is enclosed in a glass vessel, its upon thermodynamics (q.v.), for believing that apparent expansion is quite considerable. The the efficiency of the thermo-pile, as an instru- mercury-in-glass thermometer consists essenment for converting heat energy into electri- tially of four parts, these being (1) the mercal energy, can never be high. In a particular cury, (2) the spherical or cylindrical bulb of case that was investigated by Lord Rayleigh, glass which contains it, (3) the fine tube which the maximum possible efficiency was found to is attached to the bulb and which serves to make te 6 per cent; and it is not likely that an the expansion of the mercury evident, and (4) efficiency materially greater than this will ever the graduated scale which is affixed to the capilbe actually attained with a thermo-pile that lary tube and from which the indications of the is large enough and durable enough to be of instrument are read. In the manufacture of commercial value.

thermometers which are intended for accurate

ALLAN D. RISTEEN, work, the mercury is carefully purified by filDirector of Technical Research, The Travelers

tration through leather under pressure, and Insurance Company, Hartford, Conn.

by subsequent distillation, and, in many cases, THERMOGRAPH (Greek, heat-writing"),

by chemical treatment also; and immediately any form of self-registering thermometer

before the mercury is placed in the thermometer (q.v.), by which an automatic record of varia

it should be boiled so that it may be rendered

free from moisture and from air. In the mantions of temperature is kept. Many different types of thermograph have been made, of which

ufacture of the capillary stems of thermometers, the following may be especially noted: (1)

some kind of glass which experience indicates Photographic thermographs, in which the posi

to be adapted to this end is melted and the option of the mercury thread in an ordinary ther

erator takes up a ball of it on the end of his

blowpipe, blowing it out gradually and adding mometer is photographed upon a moving sensitive film, either continuously or at short in

more glass to the mass from time to time. tervals of time; the moving sensitive film be

When the ball of molten glass has thus been ing actuated by clockwork, so that the time

brought to a convenient size, a second workman

attaches his blowpipe to it also, and the two, at which any given impression was made can be nicely determined. (2) Metallic-strip thermo

still blowing, walk apart, so that the sphere of graphs, in which a recording pen is actuated by

glass is drawn out into a very long and fine

tube, which, when it has cooled, is cut into a strip of metal composed of two substances

lengths and annealed. The calibre of each of of differing expansibility, riveted or soldered

these lengths is subsequently measured under together. When a strip of this kind is heated, one of its sides expands more than the other,

the microscope, so that the instrument maker

may know how large a bulb must be attached and the result is that the strip becomes curved

to each piece, in order that the degree-spaces by an amount which serves as a measure of on the finished thermometer may be of approxithe temperature to which the strip has been

mately the desired size. The bulb of the therexposed. The pen which makes the record

mometer is usually made of a different kind of moves radially on a disc of paper which is

glass from the stem; and the process of makcaused to revolve at a steady rate by means

ing it consists simply in fusing to one end of of clockwork. (3) Electric-contact thermo

the open stem a knob of special glass and then graphs, in which a fine platinum wire is caused

blowing it to size through the stem. In therto descend, at intervals, into the open upper mometers that are to be used for accurate sciend of the capillary tube of a sensitive mercu- entific work, the bulbs should always be made rial thermometer. When the wire touches the

of one of the three kinds of glass that are remercury column, it completes an electrical cir

spectively known as "verre dur,” Jena 16" cuit, and by this means the position of the and Jena 59". The first of these is a French mercury thread in the thermometer is recorded.

glass, which has been demonstrated to be (4) The manometric thermograph, in which the peculiarly adapted for use in thermometer pressure in a closed vessel filled with a gas is taken as the index of the temperature; the

bulbs, by the elaborate experiments made at the

International Bureau of Weights and Measpressure being recorded automatically, and the ures, at Paris. The other two are made at temperature being afterward inferred from the

Jena, Germany, and have been similarly proved recorded pressure, by means of a theoretical

to be adapted for use in accurate thermometers, formula, or else by direct comparison of the in

by the experiments made at the Reichsanstalt, in strument, at different temperatures, with a Berlin. Since 1917 American glasses are manustandard thermometer. (The pressure of an

factured that are well suited to the construction isolated mass of gas of this kind is known to

of accurate thermometers; but the problem be sensibly proportional to the absolute tem- to be solved is a difficult one and it will require perature of the gas, so long as the volume is

much study and experiment. The stem and kept constant).

bulb of the thermometer being thus completed, THERMOMETER (Greek "heat-meas- the next step consists in cleaning them thorurer”), an instrument for determining the tem- oughly on the inside. For this purpose they are peratures of bodies. The general problem of washed out with hot nitric acid, with distilled thermometry is considered, in this encyclopedia, water and with ether. They are then thoroughly under the heading THERMOMETRY; and the dried, preferably by repeatedly exhausting them, present article will be devoted mainly to the while hot, with an air pump, and then filling consideration of the common mercury-in-glass them again with air that has been carefully form of the instrument, and to certain of its dried. The next operation consists in filling modifications. The mercury-in-glass thermom- the thermometer with pure mercury. In order

[blocks in formation]

to do this the bulb is heated until the air that it thermometer is complete. It may be, of course, contains is partially expelled and the open end that the thermometer is to be divided into half of the stem is then dipped beneath the mercury. degrees, or into tenths; but the operation is As the bulb cools, the air remaining within it precisely the same, in this case, as it is when contracts and mercury rises through the stem the division is to be made to degrees only. until the bulb has become partially filled; this We could evidently divide the space between operation being repeated until the bulb is full. the boiling point and the freezing point into as The instrument is next heated to a temperature many equal degrees” as we chose; for there is · considerably higher than the highest temper- no reason, in the nature of things, why a «deature to which it is to be exposed in use, the gree could not have any one size, just as well mercury that it contains becoming thereby so as any other size. It is desirable, however, to much expanded that it fills the entire stem and

have some uniform practice in this respect, and runs over at the top; and while the stem is still hence the manufacturers of thermometers infilled in this manner it is sealed off at the end

variably conform to one or the other of three by means of a blowpipe. In the higher grades standard systems. In France, and also for of thermometers, a tiny pear-shaped bulb is left scientific work in nearly every country, it is at the top of the stem, partly as a precaution

customary to follow the plan introduced by against the destruction of the thermometer in

Professor Celsius of Upsala, which consists in case it is accidentally exposed to too high a

dividing the fundamental interval between the temperature in its subsequent service and partly

two fixed points into 100 equal parts, the freezas an aid in the calibration of the stem. When such a pear-shaped bulb is provided, the

ing point being called “zero," or 0°, and the

boiling point 100°. This method of graduation stem may be sealed off at the end while the in

is known as the "Centigrade) (or “hundredternal space is exhausted by means of an air pump, instead of while it is filled with mercury;

degree”) system. For general purposes in the or the tube may be filled, above the mercury

United States and in England, it is far comcolumn, with dry nitrogen or some other inert

moner io graduate thermometers according to

the system introduced by Fahrenheit of Dantzig, gas. The glass part of the instrument having

about 1714. In this system the interval between been completed, it remains to affix the scale to the stem. In high grade thermometers, the

the freezing and boiling points is divided into scale is engraved upon the stem directly; but in

180 equal spaces, or degrees”; but the freezing the cheaper forms it is usually engraved or

point is here called 32° and the boiling point stamped upon a piece of metal or of wood, to

212o (32° + 180° = 212°). There has been which the thermometer is finally secured. Let

much discussion as to the reason that Fahrenus consider the high grade instruments first

heit had for dividing the fundamental interval and the cheaper ones afterward. Instruments

into 180 equal parts; but there can be no doubt of the former class are graduated by finding,

but that his zero point was intended to repre

sent the greatest cold that was known in his experimentally, two definite points upon the stem, corresponding to two known tempera

day, this being obtained by mixing salt and tures; the two known temperatures which are

snow. By adopting this lowest temperature as selected for this purpose being the boiling point

his zero, he probably sought to avoid the use of and freezing point of water. When these two

negative temperatures; but in these days when points are found, the space between them is temperatures several hundred degrees lower divided into a certain number of equal parts,

than his zero can be produced, the significance which are called degrees. In determining the

of the 32 is lost, and we now adhere to it simposition of the boiling point upon a thermom

ply from custom. The third thermometric syseter, the instrument is placed in steam that is

tem that has been used to a considerable extent rising from water that is boiling freely under

is that due to Réaumur. In this system the a barometric pressure equal to that which would

fundamental interval is divided into 80 equal be produced by a column of pure, ice-cold

degrees, and the freezing and boiling points mercury, 760 millimeters high, at sea-level in are marked 0° and 80°, respectively. This latitude 45°. When the mercury column in the method of graduation is extensively used in thermometer ceases rising and becomes station- Germany upon thermometers intended for ary, the point opposite which it stands is marked household purposes; but for most other purupon the stem and is called the “boiling point.” poses in that country it has given way to the If the barometric pressure under which the ex

Centigrade system. periment is performed is not identically equal After a thermometer that is to be used for to the value assumed above, allowance must be precise measurement has been made and gradmade for that fact by the aid of the experiments uated, it is subjected to certain experimental inof Regnault (or others) upon the variation of vestigations, for the purpose of ascertaining the the boiling point of water per millimeter of errors to which it may be liable. One of the change of barometric pressure. The boiling most important of these investigations relates point having been marked upon the thermom- to the "calibration error, which is due to such eter as here indicated, the instrument is then irregularities of calibre as the bore of the stem placed in a mixture of water and finely pulver- may possess. In order to determine the calibraized ice, as quickly as this can be safely done; tion errors, a thread of mercury of suitable and the point to which the mercury sinks is length is detached temporarily from the column marked and called the "freezing point.” The in the stem, by shaking the instrument. An distance, on the stem, between the boiling and expert in this kind of work can usually detach freezing points, is then marked off, by means a thread of almost any length that he pleases, of a dividing engine, into as many equal spaces whether it be long or short. The instrument is as there are degrees between the freezing and then inclined so that the detached thread may boiling points of water and (save for the affix- be brought into various positions in the stem; ing of numbers to the degree-marks) the and in certain of these positions its length is

[blocks in formation]
[ocr errors]

observed with great care. The volume of the sition of the “zero point of the thermometer, thread being constant, it is plain that its length which is very troublesome when measurements will be greater where the calibre of the tube is of the highest precision are to be made. It is small than it will be where the calibre is rela- on account of this anomalous variation in the tively large. The details of the operation of position of the zero point that the three kinds determining the calibration errors of a ther- of glass mentioned in the earlier part of this mometer are very involved; but the general article are recommended for the manufacture plan consists in observing the lengths of de- of the bulb; the variation of the zero having tached threads of mercury at different points of been studied in the case of these species of the stem and then computing from these ob- glass with great care. The phenomena as observed lengths, the relative areas of cross-sec- served in the case of «verre dur) are thus tion of the stem-calibre at various points. It is

described by Guillaume: "When a verre dur then possible to calculate a table of calibration

thermometer is quickly exposed to a temperacorrections, by the aid of which it will be easy

ture of 100° C., after having reposed for a conto correct any given reading of the instrument,

siderable time at the ordinary temperature of so as to find what reading would have been ob

the laboratory, its zero point falls with such tained if the item had been of absolutely uniform calibre throughout.

rapidity that after an exposure of one minute

at 100° C. the displacement is practically comProminent among the other sources of error, there are four that merit special attention.

plete. If the thermometer is then placed in (1) In the measurement of a temperature, the

ice-water, its zero ascends, for the first few bulb of the thermometer is supposed to be fully

moments, at the rate of about 0.001° C. per exposed to that temperature; but since the

minute; but this rate diminishes rapidly. When mercury in the stem must be seen in order to be a thermometer is maintained at a constant read, it often happens that the stem of an instru

temperature, its zero point rises little by little ment is necessarily exposed to conditions of

and the change can be traced plainly for sevtemperature that are materially different from eral years. For thermometers of verre dur, the those to which the bulb is subjected. Hence gradual rise at constant temperature amounts there is often a "stem error) to a thermometer, to about 0.001° C. per month when the therdue to the fact that the mercury thread in the mometer is two years old; and at the end of stem is colder (or hotter) than that in the bulb, four or five years the motion is found to have and, therefore, shorter (or longer) than it really diminished to about 0.0002° C. per annum.” ought to be. The magnitude of this stem error The ideal way of measuring a temperature, with will obviously vary with the conditions under a thermometer made of one of the three which the thermometer is used. It is always glasses mentioned above, is as follows: The uncertain in amount and hence it is customary, thermometer is exposed to the temperature that in well-executed scientific work, to design the is to be measured, and its zero point falls to apparatus that is to be used (including the a certain (presumably unknown) position. Afthermometer itself), with special reference to ter the instrument has been read, it is introthe desirability of keeping the stem error as duced, as quickly as is consistent with its small as possible. (2) When the barometric safety, into a mixture of water and pulverized pressure upon the bulb of the thermometer

ice. The mercury sinks at once and soon atvaries, the bulb yields elastically to these vari- tains a stable position, which, on account of ations and often to an extent quite sufficient the slowness of the change of zero with to influence the reading of the instrument by

falling temperature, is taken to be the zero an amount that cannot be neglected. The error

corresponding to the higher temperature to due to this cause can be determined and elimi

which the instrument has been previously exnated by means of the external pressure co

posed. In accordance with this plan, the efficient, which is obtained by subjecting the thermometer, at some fixed temperature, to a

temperature to be measured is found by subknown change of external pressure and noting

tracting the subsequent reading in ice-water the alteration of the reading that this variation

from the reading obtained at the temperature

to be determined. The method here outlined, of pressure produces. (3) The pressure of the

for eliminating the effect of variations in mercury upon the inner surface of the bulb may vary from several causes, one of which is the

the zero point of a thermometer, is known as

the method of movable zeros, and is now position of the thermometer itself. If the stem is in a vertical position, the bulb will be sub

adopted at practically all of the centres of acjected to a pressure due to the height of the

curate thermometry except Kew, for temperacolumn of mercury in the stem; and when the

tures between the freezing and boiling points. thermometer is horizontal, this static pressure

It is not yet possible, by any method of prowill be absent. In small thermometers the error

cedure, to determine temperatures more than a due to this cause is unimportant; but in instru

few degrees below the freezing point, or more ments of high precision, in which the stem may

than a hundred degrees (Centigrade) above the be several feet in length, it must receive due boiling point, by the aid of a mercury-in-glass consideration. The constant which is used for thermometer, with a precision comparable with correcting for this source of error and which is that which is attainable within the fundamental to be determined by experimenting with the

interval that lies between 0° C. and 100° C. thermometer in different positions but at the It is to be understood that in the foregoing same constant temperature is called the in- discussion of the errors of the mercury-in-glass ternal pressure coefficient." (4) It is found that thermometer, we have been treating of the dethe glass of which a thermometer is composed termination of temperatures to such a degree of exhibits certain anomalies in its expansion and precision that the final error is not to exceed contraction, when its temperature is altered. (say) 0.005° C. No such elaborate care is reThese result in an apparent variation in the po- quired, if the only object of the measurement is


[blocks in formation]

to determine the temperature to the nearest of temperatures approaching the freezing point degree, or half-degree.

of mercury (37.8° F. below zero) the InternaPassing now to the consideration of the tional Bureau of Weights and Measures preordinary thermometers that are used about the fers a thermometer filled with toluene to one household and by amateur meteorological ob- that is filled with alcohol; the toluene therservers, it may be pointed out, first, that in the mometer being apparently capable of yielding manufacture of a thermometer that is to be sold much more accurate results. Owing to the fact at retail for (say) 50 cents, it is not commer- that alcohol boils at a much lower temperature cially possible to engrave a special scale for than water, the alcohol thermometer can hardly each instrument. In making cheap thermom- be graduated by the method given for the mereters it is customary to stamp out the scales in cury instrument, since exposure to a temperalarge numbers and then to blow the bulb of ture of 212° F. would cause the alcohol to each instrument to such a size that the scale have a vapor pressure so high that the bulb will be as nearly as practicable adapted to the would be likely to burst. These thermometers finished thermometer. This can be done, by an are, therefore, graduated, most commonly, by experienced glass-worker, with greater accu- direct comparison with a standard mercury-inracy than might be supposed; but it is evident glass instrument. The expansion of alcohol that no high degree of precision can be at- by heat is not strictly proportional to that of tained in this way. The scale and the rest of mercury and hence if the scale of the mercury the thermometer being adapted to each other thermometer is taken as the standard, the deas nearly as is commercially practicable, the gree marks upon the alcohol thermometer will thermometer is adjusted with respect to the not be spaced at uniform intervals. These scale by exposing it to some known temperature spaces are in fact smaller at low temperatures (say 70° F.) in the vicinity of the tempera- than at higher ones, as will be seen by examintures at which it is most likely to be used and ing any good alcohol thermometer that is then securing it in such a position that the adapted for observing a considerable range of point on the stem to which the mercury rises temperature. comes opposite the proper mark on the scale. Maximum and Minimum Thermometers Such a theremometer will give readings that are thermometers which automatically record are not greatly in error at temperatures near the highest or lowest temperatures to which the one at which it is standardized; but at other they have been exposed during a given period. temperatures any two such thermometers will In the Rutherford maximum thermometer the necessarily diverge by an amount which de- capillary stem of the instrument is placed nearly pends upon the judgment and skill of the work- horizontal and as the mercury rises it pushes men who blew the bulbs and who endeavored before it a tiny index of iron or steel, placed to give them capacities adapted to the sizes of within the tube; and the index, being left at the the degrees upon their respective graduated most extreme position attained by the mercury, scales.

indicates the highest temperature to which the For further information concerning the instrument has been exposed. In the Ruthermethods that are used in precise thermometry ford minimum thermometer a similar index is consult Guillaume, 'Thermometrie de Préci- used, but the thermometric column is here comsion'; and for the historical aspect of the sub- posed of alcohol and the index lies within the ject consult H. Carrington Bolton, Evolution alcohol. When the temperature falls, the free of the Thermometer. Consult, also, Preston, end of the column of alcohol in the stem ad(Theory of Heat.)

heres to the index and drags it toward the Gas Thermometer.-A thermometer in bulb; but when the temperature rises again, which the temperature is measured by the the alcohol flows around the little index (which change of volume, or pressure, of a mass of gas does not fill the capillary tube), and so leaves enclosed in a glass envelope. The gases that it in the position to which it had been drawn are most commonly employed for this purpose at the moment when the temperature was loware air, hydrogen and nitrogen; and thermom- est. In both forms of thermometer the index eters containing these several gases are re- is returned to a suitable position for making spectively called (air thermometers," "hydro- a new observation by the aid of a small maggen thermometers” and nitrogen thermom- net. In the Negretti and Zambra maximum eters. See THERMOMETRY.

thermometer the capillary tube is partially obAlcohol Thermometer.-A thermometer in structed near the bulb so that although the which the temperature is indicated by the expan- mercury flows outward readily enough as the sion of alcohol (instead of mercury); coloring temperature rises, a fall of temperature at any matter of some kind being dissolved in the al- moment causes the mercury thread in the stem cohol, so that the column of fluid in the stem of to break at the obstruction, so that the maxithe instrument may be distinctly visible. Al- mum temperature to which the thermometer cohol has a larger coefficient of expansion than has been exposed can be read directly, in the mercury, and hence, for the same sizes of bulb usual manner. The broken thread can easily and stem, the degrees are longer upon a ther- be returned to the partially empty bulb by jarmometer containing it. Alcohol can also be ring the instrument, or by whirling it sharply used at temperatures that are low enough to in a circle. destroy an ordinary thermometer, by the freez- Clinical Thermometer.—A form of the Neing of the mercury. No great degree of preci- gretti and Zambra maximum thermometer, sion can be attained with the alcohol thermom- which is used by physicians for determining the eter, however, partly because the liquid wets the temperature of the human body. glass and thereby causes the instrument to read The graduation on these instruments is fine, too low when the temperature is falling, and so that the temperature can be read to the 10th partly for other reasons. For the measurement of a degree or so; and the entire interval cov

[ocr errors]



« PrejšnjaNaprej »