MAXIMUM ATTAINABLE TEMPERATURES.--In order to discover the maximum temperature which can be reached in or about an acetylene generator when an apparatus belonging to one of the best types is fed at a proper rate with calcium carbide in lumps of the most suitable size, the following calculation may be made. In the first place, it will be assumed that no loss of heat by radiation occurs from the walls of the generator; secondly, the small quantity of heat taken up by the calcium hydroxide produced will be ignored; and, thirdly, the specific heat of acetylene will be assumed to be 0.25, which is about its most probable value. Now, a hand-fed carbide-to-water generator will work with half a gallon of water for every 1 lb. of carbide decomposed, quantities which correspond with 320 grammes of water per 64 grammes (1 molecular weight) of carbide. Of those 320 grammes of water, 18 are chemically destroyed, leaving 302. The decomposition of 64 grammes of commercial carbide evolves 28 large calories of heat. Assuming all the heat to be absorbed by the water, 28 calories would raise 302 grammes through (28 X 1000 / 302) = 93° C., _i.e._, from 44.6° F. to the boiling-point. Assuming all the heat to be communicated to the acetylene, those 28 calories would raise the 26 grammes of gas liberated through (28 X 1000 / 26 / 0.25) = 4308° C., if that were possible. But if, as would actually be the case, the heat were distributed uniformly amongst the 302 grammes of water and the 20 grammes of acetylene, both gas and water would be raised through the same number of degrees, viz., 90.8° C. [Footnote: Let x = the number of large calories absorbed by the water; then 28 - x = those taken up by the gas. Then-- 1000x / 302 = 1000 (28 - x) / (26 X 0.25) whence x = 27.41; and 28 - x = 0.59. Therefore, for water, the rise in temperature is-- 27.41 X 1000 / 302 = 90.8° C.; and for acetylene the rise is-- 0.59 X 1000 / 26 / 0.25 = 90.8° C.] If the generator were designed on lines to satisfy the United States Fire Underwriters, it would contain 8.33 lb. of water to every 1 lb. of carbide attacked; identical calculations then showing that the original temperature of the water and gas would be raised through 53.7° C. Provided the carbide is not charged into such an apparatus in lumps of too large a size, nor at too high a rate, there will be no appreciable amount of local overheating developed; and nowhere, therefore, will the rise in temperature exceed 91° in the first instance, or 54° C. in the second. Indeed it will be considerably smaller than this, because a large proportion of the heat evolved will be lost by radiation through the generator walls, while another portion will be converted from sensible into latent heat by causing part of the water to pass off as vapour with the acetylene.