Page:
1 2 3 4 5 6 7 8
9 10
1°).—The method of finding the balance point on the bridge-wire as above described is the most sensitive and accurate for measuring a constant re sistance, as in testing the coils, or calibrating the box, or observing the fixed points of a thermometer in ice or steam. But it requires at least two accurate settings of the contact, and recording two readings of the vernier and of the corresponding deflections of the galvanometer, in addition to the coils in circuit. In many cases the highest degree of accuracy is unnecessary, and the use of the balance method is inexpedient, especially at high temperatures, where an accuracy of o.oi° C is usually ample, and the re sistance is seldom sufficiently constant to permit more exact read ings. In such cases the most convenient and rapid method of operation is to fix the bridge-wire contact at any convenient zero, set the coils to the nearest whole degree, and read the fraction of a degree by observing the deflection of the galvanometer on re versing the battery current. This takes about a tenth of the time and trouble involved in a balance reading, and is the most appro priate method to employ in reading a number of different ther mometers at regular intervals. If the thermometers are at widely different temperatures, as often happens in practice, the balancing coils as well as the thermometers must be changed at each observa tion. But since the coils required for any particular thermometer usually remain the same for an hour or so, these can be recorded before commencing observations as soon as the temperature is sufficiently steady. The change from one thermometer to the next can be effected very quickly with the box and connections as illustrated in figure 8, and it is possible to read and record the de flections at the rate of 2 or 3 a minute. The coils required for each thermometer can be readily reduced to degrees of tempera ture from a knowledge of the fundamental intervals, which seldom differ by as much as r in i,000. But the reduction of the observed deflections to the corresponding fractions of a degree, requires different factors for each thermometer in this method of opera tion, depending on the coils in circuit. The required factors are readily observed under any conditions by observing the change of deflection on reversal produced by putting the smallest coil I in or out of circuit. But it saves a good deal of trouble in practice to keep the galvanometer deflection constant at ion mm. per I° in all cases. This can be done automatically by adding a simple accessory, called a differential connector, which makes it possible to use the box for the direct observation of the difference between any two thermometers, and also permits the enclosure of the box itself in a dust-proof case.
Accurate measurements of dif ference of temperature between two thermometers are often re quired in calorimetric work, and are most readily secured by using a pair of platinum thermometers, connected on opposite sides of the bridge, so that the difference is obtained by a single reading, which is necessarily simultaneous for the two, and saves more than half the time required for reading each separately. The ar rangement of connections for reading two thermometers differen tially could easily be made on the bridge itself by omitting the resistance I and inserting a second pyrometer plug R2 in its place, side by side with the plug
belonging to the first pyrometer. The bridge reading would then give the difference
and the difference
could similarly be obtained by simply inter changing the plugs
and R2. The most accurate method of finding the difference is by observing the shift of the balance point on the bridge-wire, as in Carey Foster's method, when
and R2 are interchanged. But it is preferable in dealing with large differences to observe the coils and the galvanometer deflection, and to subtract a zero reading, taken when R2 is at the same tem perature as
This zero should be verified daily as a matter of precaution for any pair of thermometers, but is not likely to vary appreciably unless the contacts require cleaning.
Instead of making any such changes of connections on the box itself, it is more convenient in practice, for differential measure ments and similar purposes, to employ an external set of mercury cups similar to those on the top of the box, but with fewer cups. Such an arrangement may be called a differential connector. The left side of the series of cups is connected through a py rometer plug to the cups C2, P2 on the bridge, and the right side to the cups
by the same plug. The effect is the same as if the series of cups in the differential connector were interpolated between the points
and
on the bridge. But this would be inconvenient on the bridge itself, because the mechanism re quired for the resistance plugs cannot perform the manipulation of interchanging pyrometer plugs. Spaces for 4 pyrometer plugs are usually provided on the differential connector. No. i, on the left, is usually occupied by a pyrometer in steam at atmospheric pressure, serving as a convenient standard of reference in many cases. Any pyrometer R2 inserted in No. 2, can be read by dif ference from 100° C if spaces 3 and 4 are short-circuited-with resistance plugs. As a rule spaces 3 and 4 are occupied by auxiliary resistances in the form of compensated coils, made like those in the box, and enclosed in glass tubes, but provided with plug con nectors like a pyrometer.
It is a great convenience in practice, especially for differential measurements over consider able ranges of temperature, to keep the sensitivity of the galvan ometer automatically constant at a convenient figure, such as Ioo mm. per I°, in spite of any variation of resistance of the pyrometers. This requires that the measuring current should be kept constant, or that the whole resistance in circuit should remain the same, to about a half of i per cent., or less, if an accuracy of o.or° is required with a scale deflection of ioo mm. per I°. This is effected by inserting in spaces Nos. 3 and 4 on the differen tial connector one or two compensated coils having a total re sistance larger than the difference to be measured, say 19.2 ohms, equivalent to the sum of coils B and C in the box, or to 384° C on the platinum scale, if the differential thermometers
and R2 have each a fundamental interval of 5 ohms. With
R2 at the same temperature in a hypsometer, the bridge should then bal ance with B and C up in the box, and the bridge-wire contact fixed at zero. Observe the deflection, if any, produced by reversing the battery. Raise coil I, which is equivalent to 2°. Observe the deflection again. The difference of the two deflections should be 20o mm. If not, adjust the resistance of 200 ohms in the battery circuit until the desired sensitivity of ioo mm. per I° is secured. This will remain constant for a week or so if the battery is good. It will also be unaffected by any variation of R2 (or by the sub stitution of other similar thermometers in the place of
pro vided that the coils-up in the box are reduced as R2 increases so as to keep the total resistance on the left side of the bridge wire balanced against the constant resistance
on the right, within I° C. It is very easy to read to o.oi° C by this method on the galvanometer deflection, but the galvanometer deflection is less than a quarter of i per cent of the difference measured at 500° C, and it will be understood that it is essential that the resistance coils and thermometers must be most perfectly com pensated, and that the deflection must be observed by reversing the battery in order to eliminate accidental thermoelectric effects, if an accuracy of o.or° C is expected in the final result. Neglect of the last precaution alone may easily cause an error of I° if the thermometer is badly made.