The heat-transfer coefficients have been determined for five steel cylinders having fins 1.22 inches wide and the spacing between the fins ranging from 0.022 to 0.131 inch. The cylinders were tested with and without baffles in a wind tunnel; they were also tested enclosed in jackets with the cooling air supplied by a blower. A maximum heat transfer was reached at a fin space of about 0.45 inch for the cylinders tested with each of the three methods of cooling investigated. The rise in temperature of the air passing between the fins and the change in flow pattern were found to be important factors limiting the heat transfer that may be obtained by decreasing the fin space. The use of baffles for directing the air around the cylinders with closely spaced fins proved very effective in increasing the over-all heat-transfer coefficient, provided that the spacing was not appreciably less than that for maximum heat transfer.

The discharge characteristics of two similar injection valves operated by a single-cylinder fuel-injection pump were determined with an apparatus that measured the quantity of fuel discharged from each valve during every 0.5 degrees of pump rotation. It was found that similar discharges took place from the two valves at all pump speeds when the valve-opening pressures, the nozzle-orifice diameters, and the injection-tube lengths were the same for both valves. Under these conditions, the effects of changing the pump speed, the pump throttle setting, or the nozzle orifice diameter were very similar to those occurring with a single-injection valve. By a proper selection of discharge-orifice areas and valve-opening pressures it was possible to obtain a great many combinations of discharge quantities, discharge rates, and injection timings for the two valves. A series of tests using injection tubes of unequal lengths for the two valves showed that under these conditions the injection timing and the fuel quantity discharged from each valve varies widely and erratically with changes in the pump speed.

"Experimental and analytical studies were made of solid and hollow deep rectangular beams to study their lateral instability under various conditions of loading and restraint. The tests were made on bars and tubes of 17ST aluminum alloy. Failure by lateral buckling occurred only in tests on the solid beams. It was found that, within the elastic range, the test results were in agreement with the classical theory for the lateral buckling of deep beams as given by Prandtl, Mitchell, and Timoshenko" (p. 1).