The oil-tank on a Dornier-Wal stands with one of its sides outside the engine-gondola’s wall. This side is furnished with cooling-ribs for cooling off the oil. On our machine the tank was designed right into the engine’s gondola and therefore any cooling off was unnecessary. In addition to this capsules were built over the motors so that the heat from the engines could be kept in the gondolas without cooling down like the temperature outside. All the pipes were bound many times round with linen strappings. Certain pipes had the inner layer of bindings of _felt_. This provision was made both as a means of isolation from the cold and to prevent “burst pipes.” Experience here in Norway and in other lands shows us that most motor trouble on a long flight originates in one or other of the pipes. The motor conducts itself well generally. Truly I have seldom, if ever, seen a motor-construction so free from vibration as on our machines and therefore there was little possibility of burst pipes. As a safety measure, all the same, I regard such binding as necessary. To the cooling water we added 4% pure glycerine and thus had a mixture which would not have frozen before we had -17° c. and we did not have such a low temperature up in the ice regions. All the same we took the precaution of tapping the water down on to one of the petrol tanks whenever it was not necessary to be ready to start at a moment’s notice. By a special contrivance we could pump the water direct from the tank into the radiator again. We generally started the engine first, then pumped the water up. I should like to explain why. The lower part of the intake-pipe was encompassed by a water-cap through which a smaller quantity of the cooling water is led for the purpose of warming the pipe. When the propeller starts to turn, petrol begins to flow, lowering the temperature in the petrol pipe considerably below atmospheric temperature. The walls of the water-cap take on the same low temperature immediately. If the cooling mixture at this time stands at a temperature which is barely a few degrees above water’s freezing point, one runs the great risk of there being so much freezing that the exhaust of the cap will be blocked. If this occurs the cap will in one moment become a solid block of ice, causing the sides to burst in consequence. Should one, on the contrary, start the engine first and fill up, the cooling water will thus, in its passage through the cylinders, be so warmed when it reaches the cap that this calamity will be avoided.