First taking exception to the term “steam cooling” because of its association with steam heating, and suggesting “evaporative cooling” as more appropriate, the author describes the results of 7 years' experimentation with various modifications of an original system and declares that the latter has proved itself to be the most satisfactory.The problem is stated to be merely that of cooling an internal-combustion engine with boiling water, using the water as a carrier of the steam produced to transfer the very considerable quantity of heat contained in the steam to a system suitable for condensing the steam and wasting its latent heat by condensation.After detailing the many shortcomings of the conventional water-cooled system, which include the liability of the system to become stagnant because of the formation of steam-pockets and the consequent inoperativeness of the pump; the excessive size of the radiator required to meet the maximum demand while running with wide-open throttle, at maximum possible speed, in a hot climate, or at a great altitude; the unsuitability of such a radiator for normal operation during the greater part of the year; and the problem of crankcase-oil dilution due to overcooling, the advantages of evaporative cooling are outlined.Although the problem of allowing water to boil in the cylinder jackets, the steam formed to be condensed and the condensate to be returned does not seem difficult, many practical considerations must be taken into account.Experiments with numerous variations of the original system are described, such as operating at a variable or a fixed pressure; at subatmospheric, atmospheric or superatmospheric pressure; with the steam circulating downward as in ordinary radiator practice, from side to side or introduced at the bottom and allowed to flow upward; or with dry steam, wet steam or steam and water.Of these variations the most satisfactory combination is said to be that in which wet steam containing water is introduced into one side of the radiator-core, the water being allowed to traverse the lower tubes and the steam the upper tubes of the radiator. The steam in crossing the radiator is condensed and trickles down to the water and both are returned by one pump to the cylinder-block.The general principles said to apply to successful operation include: First taking exception to the term “steam cooling” because of its association with steam heating, and suggesting “evaporative cooling” as more appropriate, the author describes the results of 7 years' experimentation with various modifications of an original system and declares that the latter has proved itself to be the most satisfactory.The problem is stated to be merely that of cooling an internal-combustion engine with boiling water, using the water as a carrier of the steam produced to transfer the very considerable quantity of heat contained in the steam to a system suitable for condensing the steam and wasting its latent heat by condensation.After detailing the many shortcomings of the conventional water-cooled system, which include the liability of the system to become stagnant because of the formation of steam-pockets and the consequent inoperativeness of the pump; the excessive size of the radiator required to meet the maximum demand while running with wide-open throttle, at maximum possible speed, in a hot climate, or at a great altitude; the unsuitability of such a radiator for normal operation during the greater part of the year; and the problem of crankcase-oil dilution due to overcooling, the advantages of evaporative cooling are outlined.Although the problem of allowing water to boil in the cylinder jackets, the steam formed to be condensed and the condensate to be returned does not seem difficult, many practical considerations must be taken into account.Experiments with numerous variations of the original system are described, such as operating at a variable or a fixed pressure; at subatmospheric, atmospheric or superatmospheric pressure; with the steam circulating downward as in ordinary radiator practice, from side to side or introduced at the bottom and allowed to flow upward; or with dry steam, wet steam or steam and water.Of these variations the most satisfactory combination is said to be that in which wet steam containing water is introduced into one side of the radiator-core, the water being allowed to traverse the lower tubes and the steam the upper tubes of the radiator. The steam in crossing the radiator is condensed and trickles down to the water and both are returned by one pump to the cylinder-block.The general principles said to apply to successful operation include: