Aluminum and its alloys absorb various kinds of gas while they are in a molten state, and, when they are cast, these gases absorbed therein in excess of the normal limit while they are molten will released in the course of their solidification and cause pinholes or blowholes in their castings, and cause blister in case of sheet after annealing, thereby adversely affecting the quality of the products and increasing the risk of their rejection to the great annoyance of the makers concerned. As a consequence, numerous and varied researches hav been made in the cause of such defects and for the prevention thereof and the findings have been applied to actual factory operations with fairly commendable and efficient rusults. To-day, however, when the melting of these light metals is carried on a far larger scale than formerly, many cases, where the methods of prevention of said defects hitherto followed have failed to give satisfactory results, are reported.Confronted with such a situation, the writer has keenly felt the necessity of seriously reconsidering the adequacy or otherwise of the methods of melting aluminum and of degasification there of employed heretofore. The methods of degasification of aluminum and its alloys so far studied and put in practice aim, one and al, at removing the gas contained in them while they are molten. At the process when aluminum turn from a liquid into a solid, namely at coexistense of solid and liquid it will release the gas contained therein in excess of the normal limit, but this peculiar characteristic has never been utilized. Propory speaking, the methods of melting aluminum so far in practice represent a disregard for the aforementioned characteristic of aluminum and merely unstudied application of the metheods of melting other kinds of metals to aluminum.The writer holds the view that the release of the surplus gas in aluminum, which has generally been considered to take place at the process of the metal turning from a liquid into a solid, must occur at the process when it turns from a solid into a liquid also, viz. liquid (L)_??_s + L_??_golid (s). On the basis of this view, I have conceived an idea of liberating the released gas into the air speedily without causing it to remain in molten aluminum. In other words, the release of the surplus gas is to be so made naturaly as to take place coincidentally with only the melting process of aluminum. For an example in the case of melting a large quantity of aluminum ingot in a reverberatory furnace, for instance, the upper art of ingot is melted first and sinks down to the unmolten part beneath, mingling with and gradually melting it, but the gas released at this time will remain in the melt without being exhausted. Therefore, it is prime importance to make this gas which release in the cause of melting freely evaporate into the air. If this is possible, the degasification of molten aluminum can be achieved thereby. To put it in an extrem explanning, it is advisable to adopt a method that, in melting any aluminum ingot each molten molecule can be separated from solid as soon as it becomes melted-It can be carried out with a melting-furnace, provided with a melting support near the grate and slightly higher than the furnace-bed with an inclination to the holding-tray, and so made as, when an aluminum ingot to be melted is placed there and melted, the molten aluminum will move, bit by bit, on to the bottom of the tray without being over-heated. This method is already being actually used with excellent results. in practical production. Writer named it candle-like meltingThe above stated research was made principally with reference to the degasification of surplus gas in molten aluminum.