reduced this quantity to an average of 141.5 parts for one hundred parts raw paper. The weight of sanded paper is very variable, as it depends altogether upon the size of the sand grains. It may be stated generally that the weight of the sand is as large as that of the tarred paper.

The kinds of roofing paper saturated with other additions besides coal tar form a separate class, in order to neutralize the defects inherent in coal tar. These additions were originally for the purpose of thickening the paper and making it stiffer and drier. The most ordinary and cheapest thickener was the coal pitch. Although the resinous substances are increased thereby, still the light tar oils remain to evaporate, and the paper prepared with such a substance readily becomes hard and brittle. A better addition is the natural asphaltum, because it resists better the destroying influence of the decomposition process, and also, to a certain degree, protects the coal tar in which it is dissolved. The addition of natural asphaltum doubtless caused the name of "asphaltum roofing paper." Resin, sulphur, wood tar and other substances were also used as additions; each manufacturer kept his method secret, however, and simply pointed out by high sounding title in what manner his paper was composed. In most cases, however, this appellation was applied to the ordinary tar paper; the impregnating substance was mixed only with coal pitch, or else a roofing paper saturated with distilled tar. The costly additions, by the use of which a high grade of roofing paper can doubtless be produced, largely increased its price, and on account of the constant fall of prices of the article, its use became rather one of those things "more honored in the breach than in the observance," and was dispensed with whenever practicable. The crude paper is the foundation of the roofing paper. The qualities of a good, unadulterated paper have already been stated. At times, the crude paper contains too many earthy ingredients which impair the cohesion of the felted fibrous substance, and which especially the carbonate of lime is very injurious, as it readily effects the decomposition of the coal tar. The percentage of wool, upon which the durability of the paper depends very largely, is very small in some of the paper found in the market. In place of woolen rags, cheap substitutes have been used, such as waste, which contains vegetable fibers. Since this cannot resist the decomposition process for any length of time, it is evident that the roofing paper which contains a noticeable quantity of vegetable fibers cannot be very durable. To judge from the endeavors made to improve the coal tar, it may be concluded that this material does not fully comply with its function of making the roofing paper perfectly and durably waterproof. The coal tar, be it either crude or distilled, is not a perfect impregnating material, and the roofing paper, saturated with it, possesses several defects. Let us in the following try to ascertain their shortcomings, and then express our idea in what manner the roofing paper may be improved. It was previously mentioned that every tar roofing paper will, after a greater or smaller lapse of time, assume a dry, porous, friable condition, caused by the volatilization of a part of the constituents of the tar. This alteration is materially assisted by the oxygen of the air, which causes the latter to become resinous and exerts a chemical influence upon them. By the volatilization of the lighter tar oils, pores are generated between the fibers of the roofing paper, into which the air and humidity penetrate. In consequence of the greatly enlarged surface, not only the solid ingredients of the tar, which still remain unaltered, are exposed to the action of the oxygen, but also the fibers of the roofing paper are exposed to decomposition. How destructive the alternating influence of the oxygen and the atmospheric precipitations are for the roofing paper will be shown by the following results of tests. It will have been observed that the rain water running from an old paper roof, especially after dry weather, has a yellowish, sometimes a brown yellow color. The supposition that this colored rain water might contain decomposition products of the roofing paper readily prompted itself, and it has been collected and analyzed at different seasons of the year. After a period of several weeks of fair weather during the summer, rain fell, and the sample of water running from a roof was caught and evaporated; the residue when dried weighed 1.68 grammes. It was of a brownish black color, fusible in heat and readily soluble, with a yellow brown color in water. The dark brown substance readily dissolved in ammonia, alcohol, dilute acid, hydrochloric acid, sulphuric acid, and decomposed in nitric acid, but did not dissolve in benzine or fat oil. After several days' rain during the summer, a quantity of the water was caught, evaporated, and the residue dried. Its characteristics were similar to those above mentioned. By an experiment instituted in water under conditions similar to the first mentioned, the dry brown substance weighed 71 grammes. It possessed the same characteristics. In the solution effected with water containing some aqua ammonia of the brown substance, a white precipitate of oxalate of lime occurred when an oxalate of ammonia solution was added, but the brown substance remained in solution. A further precipitation of oxalate of lime was produced by a solution of oxalic acid, but the brown organic substance remained in solution. This organic substance being liberated from the lime was evaporated, and left a dry, resinous, fusible brownish black substance, which also dissolved readily in water. It will be seen from these trials that the substance obtained from the rain water running from a paper roof is a combination of an organic acid with lime, which readily dissolves in water, and that also the free organic acid combined with the lime dissolves easily in water.

The question concerning the origin of this organic substance or its combination with lime can only be answered in one way, viz., that it must have been washed by the rain water out of the paper. But since such a solid substance, easily soluble in water, is contained neither in the fresh roofing paper nor in the coal tar, the only deduction is that it must have arisen by the decomposition of the tar, in consequence of the operation of the oxygen. The lime comes from the coating substance of the roof, for which tar mixed with coal pitch was used. The latter was fused with carbonate of lime. These analyses furthermore show that the formation of the organic acid easily soluble in water depends upon the season; and that a larger quantity of it is generated in warm, sunny weather than in cold, without sunshine. This peculiarity of the solid, resinous constituents of the coal tar, to be by the operation of the atmospheric oxygen altered into such products that are readily soluble in water, makes the tar very unsuitable as a saturative substance for a roofing paper. How rapidly a paper roof can be ruined by the generation of this injurious organic acid will be seen from the following calculation: Let us suppose that an average of 132 gallons of rain water falls upon ten square feet roof surface per year, and that the arithmetical mean 0.932 of the largest (1.680) and smallest number (0.184) be the quantity of the soluble brown substance which on an average is dissolved in one quart of rain water; hence from ten square feet of roof surface are rinsed away with the rain water per year 466 grammes of the soluble decomposition products of the tar. The oxidation process will not always occur as intensely as by a paper roof, ten years old and painted two years ago, which instigated above described experiment. As long as the roofing paper is fresh and less porous, especially if the occurring pores are filled and closed again by repeated coatings, oxidation will take place far less rapidly. Besides this, the protective coating applied to the roof surface is exposed most to this oxidation process. Even by assuming this