water, and that gas absolutely deprived, as far as possible, of aqueous vapour does not deposit naphthalene under the ordinary conditions of temperature and pressure."

Friction, even when the gas is at ordinary temperatures, is one of the causes of the deposition of naphthalene, while sudden changes of temperature intensifies the evil. It has frequently been noticed that condensation of naphthalene, and its deposition as crystals in the pipes, are promoted by and usually occur where there is any roughness or burr on the inner surface of the pipe, on the well-known principle that crystallization is promoted by the presence of a solid body in a crystallizable liquid, or to a lesser extent where there is a sharp bend in the pipes, which, by producing friction on the gas, acts somewhat in the same manner as a burr or roughness; further, most of all the deposit occurs at places where the gas is subjected to a sudden lowering of temperature.

The circumstances under which a deposit of naphthalene most frequently takes place are when the pipes are subjected to great atmospheric changes, or when a pipe passes into and out of the ground. As instances of the first class, we may cite the effect of a shower of rain in summer or of a frost in winter upon the street lamps, and also the frequent stoppages which take place in pipes which are led across bridges, and which are usually exposed or inadequately protected from the action of the atmosphere. As an instance of the second class, the following remarks of Mr. G. Anderson will serve as a typical illustration :

"In one case which came under my notice the inlet of a gas-holder was filled with naphthalene. The inlet pipe went down into the ground, and at a certain level (about ten feet under ground) there was water in the soil. When the cap was taken off, in order to plunge' the pipe, it was found that the deposition commenced exactly at the water's edge, and that down to that point the pipe was clean."

6

,

From the foregoing it will be seen how necessary it is to have all pipes free from burrs and projections, and in laying mains to avoid sharp turns and angles. Further, it is an advantage to protect the exposed parts of mains and service-pipes, especially the pipes in the works at the points where they pass into wet soil, as at the outlets and inlets of gasholders, and wherever else the pipes pass from one medium into another which is colder. At such points, as also where the pipes are led across bridges, it will be found highly advantageous to jacket the pipes, or surround them with cheap non-conducting material.

Naphthalene deposits may be cleared from service pipes by means of the force-pump, or by pouring naphtha into them, naphthalene being soluble in naphtha.

CHAPTER XVII.

THE CONSTRUCTION OF GAS METERS.

`HERE are two kinds of gas meters in use, the "wet"

TH

meter and the "dry" meter. The former derives its name from the fact that it requires a quantity of water in order to work it, while the latter works without water, being complete in itself. Both descriptions of meter are employed for the registration of gas on consumers' premises, but the gas made on the works is always registered by a "wet" meter, known as the Station Meter.

FIG. 53.

The wet meter may be said to consist of a cast-iron outer case containing a cylindrical vessel of tinned iron, known as the measuring-wheel or drum, which is capable of freely revolving upon an axle or shaft in a horizontal position in suitable bearings. The drum is immersed in water, which fills the outer case to a predetermined level. It is of different diameters, according to the size of the meter, and is divided into four longitudinal compartments, arranged somewhat in the form of four blades of an Archimedean screw, the inlet end being isolated from the outlet side of the drum by the contrivance known as 'the spout

66

and hollow cover," shown in fig. 53, which is a section of a wet station meter; from which it will be seen that the gas enters at one end of the drum and leaves it at the other, but, by reason of the manner in which the partitions are arranged, there is not a clear way through the drum, the opening at one end of a chamber being above the waterline at the same time that the corresponding opening at the other end is below it. Two of the compartments are constantly above the water-line, the one filling with gas, the other discharging.

The openings at each end of the chambers through which the gas passes are long narrow passages, shown in section in the figure, and in elevation, are of a triangular form, and are known as "hoods." The gas entering the measuring drum presses against the under-side of the longitudinal division, and turns the drum round for a quarter of a revolution, expelling in the process the gas contained in that chamber which has immediately preceded it, and so on. Each chamber in rotation, as it fills with gas, turns the drum round a quarter of a revolution, and expels the gas from the chamber which precedes it.

The cubical space in each chamber when out of the water and full of gas is known. Four times this space represents the available capacity of the drum, and, consequently, the quantity of gas that has been contained by it in each revolution.

The spindle or shaft of the drum is in connection with a train of wheels, the latter being provided with pointers, which traverse the meter dials, and in this way the number of revolutions made by the drum is recorded, and, consequently, the amount of gas which has passed through the instrument is correctly registered. Fig. 54 is a section of a meter, direct through the centre of the drum.

The superiority claimed for wet over dry meters as accurate measurers of gas, lies in the fact that for a flexible

measuring chamber one of rigid construction is substituted, and for dry slide-valves, the hydraulic seals of the drum partition, from which it will be seen that the accuracy of the registration of the wet meter largely depends upon maintaining the water in the meter at the correct level; for if the water-line be low, the available gas space in the measuring drum will be more than it should be, and the meter will register slow, i.e., it will register a smaller quantity of gas than that which has actually passed through, while if the water line is high, the gas space in the measuring drum will be cor

respondingly reduced, and the
meter will register a quantity of
gas in excess of that actually
passed through. In this case the
meter will register fast. The best
way of maintaining a constant
water level is to place the outlet
of the water overflow pipe at the
proper level, and then to have a
small stream of water constantly
flowing into the meter, but this
method of regulating the water
level is only applicable to station meters.

FIG. 54.

In order to insure the water-line of consumers' meters. being always maintained at the proper level, various devices have been introduced under the name of compensating meters. Most meters of this class have a reservoir of water within the meter case, the water being distinct from that in which the measuring drum revolves; this reservoir of water serves to compensate for any diminution of water in the body of the meter by evaporation, or otherwise; the water is transferred to the body of the meter by various automatic contrivances, such as by the agency of spoons, scoops, etc., which, being connected to the spindle of the