and not acid in solution the various solutions were carefully tested with litmus paper for acid without detecting the minutest trace. Then for the purpose of testing the litmus paper a single drop of hydrochloric acid placed in the solution turned the blue litmus to a light pink, proving conclusively that the test paper was of proper strength. Then consider that all the time the engine is in operation the temperature is nearly to the boiling point of the solutions, in some cases more, and it will be seen that the degree of electrical activity is considerably increased. The cellular cooler is composed of innumerable soldered joints and at every one of these there will be a certain amount of electrical action, which in the aggregate will amount to a considerable current. At various other points of the cooling system, wherever there is two unlike metals in combination, we have other small currents, which decompose their quota of metal and assist in filling the system with sediment and foreign matter, not to mention the salt crystals which will be formed as the solution evaporates. The writer does not claim that the test showed absolute results, but they demonstrated that without doubt electrical action does exist when solutions of calcium chloride or any other salt are used to prevent freezing. The Best Mixture.-Plain water and alcohol solutions would be the best were it not for the ease with which such compounds boil and the rapidity with which they evaporate. We have seen that the objections advanced against calcium chloride solution have ample foundation and that such compounds are not suitable for use, the chief advantage, that of cheapness, having been eliminated by the reduction in the price of denatured alcohol. The addition of a little glycerine to an alcohol and water solution reduces liability of evaporation, and when used in such quantities it has no injurious effect to speak of on rubber hose. The tables show the combinations and their freezing points and the proper proportions of the mixtures used must, of course, be governed by conditions of locality, but it is better to be safe than sorry, and make the solutions strong enough for the extremes that may be expected. The writer has used both alcohol and water, and glycerine, alcohol and water solutions, with good results, though considerable trouble has already been expe rienced when saline solutions were employed. Oils of various kinds have been recommended, these being of the character often used in lubricating icemaking machinery, and made especially to withstand low temperatures. Such oils will not absorb heat as well as water and should be used only where exceptionally good methods of cooling are provided, such as a large radiator, all metal pip ing and positive pump. This oil will attack rubber hose, however, and it would seem, all things considered, alcohol solutions are preferable to all others. The following tabulations give the relative values of solutions commonly employed: WATER, ALCOHOL AND GLYCERINE SOLUTIONS Water 85%, Alcohol-Glycerine 15%....Freezing point, 20° F. EXTRACTS FROM A PAPER READ BY EDW. H. BARRIER BEFORE THE AMERICAN SOCIETY OF MECHANICAL ENGINEERS. Extinguishing Fires in Volatile Liquids.-The extinguishing of fires in oils, gasoline and in most of the volatile liquids has always been a difficult problem and where fires of this kind occur the results are frequently very disastrous. Our most common extinguishing agent, water, works rather unsatisfactorily upon the majority of such fires, but it is still the only one available where heroic measures are required. Comparatively recently, however, there have been two or three other materials introduced for use as extinguishers which have shown some promise for dealing with these fires, and it is the writer's purpose to discuss these materials and the conditions under which they prove the most efficient. Not all fires in volatile liquids are difficult to handle with water. When the liquid is miscible with water this extinguishing agent can be successfully used. Examples of this kind are denatured alcohol, wood alcohol, grain alcohol, acetone, etc. Where the liquid is not miscible with water little or no effect is produced except to wash the burning liquid out of the building where it may be completely consumed, or, if the quantity of oil is small, possibly to extinguish the fire by the brute cooling effect of a large quantity of water sprayed upon the fire. Soda and acid extinguishers are somewhat Fig. 466.-Charts Showing Freezing Points of Various Mixtures of Calcium Chloride, Alcohol, etc., and Water. more effective than pure water, but even they fail under most conditions. The various grenades containing salt solutions which were formerly extensively exploited are of course practically worthless. The only principles that can be made use of in extinguishing fires in volatile oils are (a) to form a blanket either of gas or of solid material over the burning liquid which will exclude the oxygen of the air or (b) to dilute the burning liquid with a non-inflammable extinguishing agent which is miscible with it. Sawdust and Bicarbonate of Soda.-To the blanketing type of extinguishers belongs sawdust. Paradoxical as it may seem, ordinary sawdust is an excellent extinguishing agent for certain volatile liquids, especially those of a viscous nature. A considerable number of experiments were conducted in the fall of 1912 by the inspection department of the Associated Factory Mutual Fire Insurance Companies, in the extinguishing of fires in lacquer and gasoline in tanks with sawdust, and the results were surprisingly satisfactory. The liquids were placed in three tanks 30 inches long, 12 inches wide and 16 inches deep; 48 inches long, 14 inches wide and 16 inches deep; and 60 inches long, 30 inches wide and 16 inches deep. The sawdust was applied with a long-handled, light but substantially built snow shovel having a blade of considerable area. In every case the fires were extinguished readily, especially in the two smaller tanks which were about as large as any ordinarily employed for lacquer in manufacturing establishments. The efficiency of the sawdust is undoubtedly due to its blanketing action in floating for a time upon the surface of the liquid and excluding the oxygen of the air. Its efficiency is greater on viscous liquids than thin liquids, since it floats more readily on the former than on the latter. The sawdust itself is not easily ignited, and when it does become ignited it burns without flame. The burning embers have not a sufficiently high temperature to reignite the liquid. The character of the sawdust, whether from soft wood or hard wood, appears to be of little or no importance, and the amount of moisture contained in it is apparently not a factor, so that the drying out of sawdust when kept in manufacturing establishments for a time would not affect the efficiency. It was found that the admixture of sodium bicarbonate greatly increased the efficiency |