We may suppose the air in the neighbourhood of the depositing surface to be reduced to such a state that it will deposit moisture, by altering its temperature merely, without altering its pressure, and accordingly without altering the pressure of aqueous vapour contained in it. We have, therefore, only to look out in a table the saturation pressure of aqueous vapour at the temperatnre of the dew-point and we obtain at once the quantity desired, viz. the pressure of vapour in the air before it was cooled.

We may compare the result thus obtained with that given by the wet and dry bulb thermometers. In this case the observation consists simply in reading the temperature of the air t, and the temperature of a thermometer whose bulb is covered with muslin, which is kept constantly moist by means of a wick leading from a supply of water. The wick and muslin must have been previously boiled in a dilute solution of an alkali and well washed before being mounted, as otherwise they rapidly lose the power of keeping up a supply of moisture from the vessel.

The pressure e" of aqueous vapour can be deduced from the observations of t and t by Regnault's formula' (available when is higher than the freezing point)

e"=e0009739t' (t-t')—5941(t−t)

-0008(t-t)(b-755)

where e' is the saturation pressure of aqueous vapour at the temperature, and b is the barometric height in millimetres.

Experiments.-Determine the dew-point and the pressure of aqueous vapour by Dines's Hygrometer, and also by the wet and dry bulb thermometer.

The reduction of observations with the wet and dry bulb thermometers is generally effected by means of tables, a set of which is issued by the Meteorological Office. The formula here quoted is Regnault's formula (Ann. de Chimie, 1845) as modified by Jelinek. See Lupton, table 35.

Regnault's hygrometer consists of a brightly polished thimble of very thin silver, forming the continuation of a short glass tube to which the silver thimble is attached by plaster of paris or some other cement not acted upon by ether. Through a cork fitting tightly into the top of the glass tube pass two narrow tubes of glass, one (A) going to the bottom of the thimble, the other (B) opening at the top of the vessel just below the cork; also a sensitive thermometer so placed that when the cork is in position, the bulb (which should be a small one) is close to the bottom of the thimble.

If, then, ether be poured into the thimble until it more than covers the thermometer bulb, air can be made to bubble through the liquid either by blowing into the tube (A) or sucking air through (B) by means of an aspirating pump of any sort. The passage of the air through the ether causes it to evaporate and the temperature of the liquid to fall in consequence, while the bubbling ensures the mixing of the different layers of liquid, and therefore very approxi mately, at any rate, a uniform temperature of silver, ether, and thermometer. The passage of air is continued until a deposit of dew is seen on the silver, which shews that the temperature of the silver is below the dew-point. The thermometer is then read, and the temperature of the apparatus allowed to rise until the deposit of moisture has completely disappeared, when the thermometer is again read. The temperature is now above that of the dew-point, and the

mean of the two readings so obtained may be taken as the temperature of the dew-point, provided that there is no more difference than two or three tenths of a degree centigrade between them.

In case the difference between the temperatures of appearance and disappearance is a large one, the method of proceeding suggested by Regnault should be adopted. The first observation will probably have given the temperature of dew appearance within a degree; say the observation was 5°; pass air again through the ether and watch the thermometer, and stop when a temperature of 6° is shewn. Then aspirate slowly, watching the thermometer all the time. Stop as each fifth of a degree is passed to ascertain if there be a deposit of dew. As soon as such a deposit is formed, stop aspirating, and the deposit will probably disappear before the temperature has risen o°2, and we thus obtain the dew-point correct to o°*1.

The thermometer should be read by means of a telescope some 6 feet away from the instrument, and every care should be taken to prevent the presence of the observer producing a direct effect upon the apparatus.

It is sometimes very difficult, and never very easy, to be certain whether or not there is a deposit of dew on the silver, the difficulty varying with different states of the light. It is generally best to have a uniform light-grey background of paper or cloth, but no very definite rule can be given, practice being the only satisfactory guide in the matter.

A modification of Regnault's apparatus by M. Alluard, in which the silver thimble is replaced by a rectangular brass box, one face of which is surrounded by a brass plate, is a more convenient instrument; the contrast between the two polished surfaces, one of which may be covered with the dew while the other does not vary, enables the appearance of the deposit to be judged with greater facility. The method of using the instrument is the same as for Regnault's.

The dew-point being ascertained as described, the

pressure of aqueous vapour corresponding to the temperature of the dew-point is given in the table of pressures based on Regnault's experiments,' since at the dew point the air is saturated with vapour. We have already seen (p. 301) that we may take the saturation pressure of vapour at the dewpoint as representing the actual pressure of aqueous vapour at the time of the experiment.

Experiment.-Determine the dew-point by Regnault's Hygrometer, and deduce the pressure of aqueous vapour. Calculate also the density of air in the laboratory at the time of observation.

THE first experiments to be performed in optics will be on the comparison of the intensities of two sources of light. We shall describe two simple methods for this, Bunsen's and Rumford's, both founded on the law that the intensity of the illumination from a given point varies directly as the cosine of the angle of incidence upon the illuminated surface and inversely as the square of the distance of the surface from the luminous point. So that if I, I' be the illuminating powers of two sources distant r, respectively from a given surface, on which the light from each falls at the same angle, the illumination from the two will be respectively I/r and I'/', and if these are equal we have

so that by measuring the distances r and we can find the ratio of 1 to I'.

Lupton's Tables, No. 34

Now this supposes that it is possible to make the illumination from each source of light the same by varying the distances of the two sources from the screen. As a matter of fact, this is not necessarily the case; in performing the experiment we compare the two illuminations by the effect produced on the eye, and that effect depends partly on the quantity of energy in the beam of light reaching the eye, partly on the nature of the rays of which that beam is composed. To define the intensity of a beam, we require to know, not merely the quantity of light in it, but also how that light is distributed among the differently coloured rays of which the beain is composed. Any given source emits rays, probably of an infinite number of different colours. The effect produced on the eye depends on the proportion in which these different colours are mixed. If they are mixed in different proportions in the two beams we are considering, it will be impossible for the effect of each of the two, in illuminating a given surface, ever to appear the same to the eye.

This constitutes the great difficulty of all simple photometric measurements. Two different sources of light, a gas flame and a candle for example, emit differently coloured rays in different proportions; the gas light contains more blue than the candle for the same total quantity of light, and so of the two spaces on which the illumination is to be the same, the one will appear bluish, the other reddish.

.

Strictly, then, two different sources of light can only be compared by the use of a spectro-photometer, an instrument which forms the light from each source into a spectrum and then enables the observer to compare the intensity of the two for the different parts of the spectrum. One such instrument will be described in a subsequent section (§ 67).

45. Bunsen's Photometer.

Two standard sperm candles (see p. 23) are used as the standard of comparison. These are suspended from the arm