There are two methods of circulating the water through the cylinder jackets and radiator. The most common method consists of the introduction of a pump in the lower portion of the circulating system. In the case of automobile motors, this pump is driven by gears connected with the crank shaft of the engine. Such a pump will be either of the gear or centrifugal type, and will suck the cooled water from the lower portion of the radiator, and force it through the jackets. The second method is known as the thermo-syphon system because the circulation is automatic and depends upon the cooling of the water in the radiator. When the cooled water sinks, a syphon action is formed that tends to draw the hot water from the cylinder jackets, and the automatic circulation will thus continue as long as the successive heating and cooling take place.

Inasmuch as the pump is driven by the crank shaft of the engine, its speed will be proportional to that of the motor. The same holds true of the fan that serves to draw the air through the radiator. It will thus be seen that both the water and the air are forced at a more rapid rate when the motor runs at high speed, and that therefore the extra heat generated by the more frequent explosions in the cylinders will be counteracted to a certain extent. The increased number of explosions and the higher speed at which the fan turns also cause quicker heating and cooling of the water by the thermo-syphon system, thus forming a more rapid circulation. Inasmuch as the force exerted upon the water by its cooling and heating is not as great as that formed by a high-speed and efficient pump, the pipes and connections of the thermo-syphon system must be of ample size in order to keep the resistance to the passage of the water as low as possible. Care must also be taken in the design of this system so to construct and connect the pipes and jackets that the hot water will be allowed to rise and the cool to descend, and thus to make possible the syphon conditions on which principle the circulation is based.

The ability of the radiator to carry off the heat from the water depends upon the rapidity with which the air passes through the passages provided for the purpose. The amount of air passing through is determined by the speed of the suction fan and the rapidity of travel of the car itself against the wind. It has been shown that, when the motor runs at a high number of revolutions, the fan turns faster and the rapidity of circulation is increased. But if the car itself does not increase its speed in proportion to the higher revolutions of the motor, the maximum amount of air will not be forced through the radiator passages, and the excess heat will not be carried off entirely from the cylinders. This is a condition that prevails when the motor is run on low gear. The speed of the motor is increased, while that of the car is reduced; additional heat is generated in the cylinders, but the speed of the air is not increased in proportion. Therefore a motor that is driven a long distance on the low gear will have a tendency to overheat.

Water under atmospheric pressure cannot be brought to a temperature above 212 degrees Fahrenheit without being converted into steam. Therefore, when the heat from a water-cooled motor cannot be carried away sufficiently fast, the water in the circulating system will begin to boil. As long as water remains in the jackets, the temperature of these spaces cannot well rise above 212 degrees, and consequently there is small danger that a water-cooled motor will become overheated to the point at which the pistons will "seize" in the cylinders. The moment the water in the circulating system begins to boil, however, exceedingly rapid evaporation naturally takes place, and the water will soon entirely disappear in the form of steam and vapor. To run the motor under these conditions will mean that pistons and rings will soon become stuck in their cylinders, although liberal quantities of oil will sometimes delay this inevitable result.

But even when the cooling water is not brought to the boiling point there is a vapor that is constantly dispelled from it whenever its temperature is brought above that of the air. The water system of an automobile must therefore be replenished at irregular intervals, depending upon the amount and nature of the running to which the car has been subjected. The older cars were provided with an extra water tank, generally located under the seat, and connected directly with the water jackets and the radiator. The usual water-cooling system of the present-day car, however, is self-contained—that is, there is no separate tank for the storage of the water. The water is poured into the top of the radiator, and from this high point it reaches every part of the circulating system. Whenever the radiator will accommodate a couple of quarts, or more, it is well to fill it, for too much water cannot be used on the modern design of cooling system. It is true that a motor runs at its highest efficiency when its temperature is as great as that at which proper lubrication of the pistons can be obtained—for a gasoline engine is a "heat engine," and the greater its unnecessary heat losses, the less will be the power developed by it. But a motor cannot be kept at the proper temperature by reducing the amount of cooling water in its circulating system. The best method is to lessen the rapidity with which the water is cooled, and this may be accomplished by placing a leather flap, a cardboard, or other obstruction over a portion of the radiator to reduce the number of openings through which the air may pass. It should only be necessary to do this in the coldest weather, however, for the cooling system of every motor is designed to maintain the proper temperature on all except the hottest or coldest days.

It has been stated in a preceding paragraph that continued running on the low gear is the most frequent cause of overheating a motor. This is true, but it is not the only cause. Obstructions in the circulating system that reduce the flow of water will have this effect, as will also deposits on the interior of the cylinders that serve to prevent the proper transfer of heat to the water in the jacket spaces. Removal of the carbon will remedy the latter trouble, but to clear out the circulating system is more or less of a complicated matter. Stoppage in the pipes or radiator cells may be caused by a lime deposit from "hard" water that may have been used in the circulating system. There are preparations intended to remove this deposit, but such should not be used without first advising with the maker of the car or an experienced repair man. A series of battered cells in the radiator may reduce the number of cooling spaces that should be traversed by the water, and thus the hot water cannot be distributed over as great an air area as is necessary to maintain the motor at the proper temperature. Such a condition will be apparent from a marked difference in temperature between the affected portion of the radiator and the remainder. If a deposit has been formed on a certain series of cells, or if they have been obstructed in any other manner, the hot water cannot circulate through this section of the radiator, and it will remain comparatively cool.

Water is a liquid that remains in its fluid stage only through a temperature range of 180 degrees—at atmospheric pressure. At 212 degrees it boils and turns to vapor, while at 32 degrees it freezes and becomes a solid. In neither of these stages does it form a desirable cooling medium for a gasoline motor. Of the two, however, its solid stage is the more harmful to the motor. Not only will it cease to flow when it becomes ice, but the expansion of the water during the formation of the solid is liable to burst its retainer—whether it be the cells of the radiator, the pump, pipes, or even the cylinder walls themselves. It is the radiator that is the most liable to suffer from such a cause, however, for each cell contains so small an amount of water that the liquid will be brought to the freezing point before the larger volume in the jacket spaces approaches this temperature. Of course the water will be kept well above the freezing point when the motor is running, and it is only when the machine has stood idle for several hours that care must be taken to prevent the formation of ice in the circulating system.

Aside from keeping the car in a warm place whenever the motor is to be at rest more than two hours, there is only one method of preventing the cooling water from freezing, and that is by the introduction of some chemical that lowers the point at which the liquid will turn to a solid. There are several ingenious heaters available that are attached to the circulating pipes and that serve to keep all of the jacket water warm; the use of these producing the same conditions as though the car were kept in an artificially-heated garage.

One of the most common liquids used in the cooling water to prevent freezing is alcohol. If equal parts of wood alcohol and water are used in the cooling system, the resulting mixture will not freeze until it reaches a temperature colder than 25 degrees below zero. A weaker mixture—one having 25 per cent. of wood alcohol—will freeze at about zero, and it therefore depends upon the prevailing cold-weather temperature as to the proper proportion that should be used. It must be remembered that the boiling point of alcohol is much lower than is that of water, and that therefore a mixture that will not freeze in exceedingly cold weather is liable to boil away on the first moderate day on which the car is run. The above-mentioned 50 per cent. mixture of wood alcohol and water will boil at 135 degrees, while the 25 per cent. solution will withstand a temperature 40 degrees higher before it is transformed into vapor. As the lower temperature will be reached easily if the motor is run for some time in comparatively moderate weather, it will be seen that the stronger mixture should be used only where winters are very severe. It must also be borne in mind that, as alcohol boils more readily than does water, it follows that it will evaporate more easily, as well. Therefore, in order to maintain a uniform proportion of wood alcohol to water, the former should be replenished more often than is the latter.

Glycerine is another substance that is often mixed with the cooling water to prevent the latter from freezing. A 50 per cent. mixture of this and water has a freezing point of about zero, or slightly lower, and boils at practically the same temperature as water—210 degrees. Combinations of wood alcohol and glycerine may be used—equal parts of each being the usual proportion—and thus various freezing and boiling points may be obtained.