approximate somewhere between 50° and 55° Fahr. during the entire year. The above statement is based upon the observed temperature of perennial springs in this vicinity.

"From the foregoing it follows that if the air in the cellar becomes warmer than the air in the underground flue, it will rise through the ventilating flues, and the colder air will flow in from the supply flue, as desired. The temperature of the cellar room can thus be approximately controlled to at least the neighborhood of 55° to 60° Fahr. The construction of the cellar is shown somewhat in detail in the drawings below. These figures are not intended as working drawings, but serve to bring out the essential ideas and plan of the structure sufficiently to enable any mechanic to carry them out on larger or smaller scale to suit the needs of the builder. Fig. 113 is a

longitudinal section through the cellar room, and shows also a side elevation of the storeroom above. The two ventilators a a rise through the storeroom, and are six inches in diameter by fifteen feet long, thus insuring good draft. The air flue b

Virginia Storage House.

455

enters under the foundation and discharges fresh air into the cellar room near the center. This flue is six inches in diameter, and theoretically should be extended far enough along the hillside to admit of tempering the air to the temperature of the

surrounding earth while passing through it. The cellar built here has an air flue only one hundred and fifty feet long, and we have never been able to cool the air down below 60° Fahr. when the temperature of the outside air is above 70° Fahr.

"The dotted line h h shows the contour of the hillside, and the line shows the entry-way into the cellar. The entry-way should preferably be on the north side of the structure, and should be closed in by a vestibule, so as to protect the cellarway from storm, and to prevent influence of outside temperature upon the atmosphere in the cellar.

"The roof structure of vestibule is shown at d, and one of the side walls of the entry-way at e. The floor of the cellar ff pitches slightly to the mouth of the air flue b, which serves as a drain pipe when one is needed. The cellar floor is made of broken stone and cement, and successfully checks the upward rise of ground water. The bins c c and c' on one side of the cellar room are shown, with dimensions. They are constructed of two by four scantling and one inch oak boards. At c' the facing on lower bin is shown in position. When we desire to fill the bins additional facing boards are used. The letters k k

at lower ends of ventilating flues indicate position of sliding dampers, by means of which the flow of air from the cellar is effectually controlled. With the short flue used in the cellar here, we find that when the mercury remains below 20° Fahr. for any length of time the cellar will be reduced below freezing, unless the ventilators are closed.

"The dimensions of the ground plan of the cellar are shown in Fig. 114. The letters so far as used always indicate the same part of the structure in the several figures.

This figure shows the walls to be constructed of stone, which is unquestionably the proper material. In our building, which was con structed to test the practicability of an idea new to this class of structures, expense was avoided so far as possible, and the walls are built of wood. The framing of the walls is constructed of two by four scantling framed into sills laid in broken stone and cement. The corner posts are four by four scantling. This frame is covered outside by a double sheathing of inch oak plank. The first course was put on diagonal and covered with strong builder's paper, and over this a perpendicular course of sheathing was put on. The whole structure was then literally soaked with crude petroleum, and the earth rammed in tight around the cellar story and banked up, so as to carry surface water away from the walls. Inside the walls were covered with inch oak boards, and the bins constructed as indicated in the drawings. The entry-way to the cellar is wide enough to admit of backing a horse-cart or wagon down into the cellar, so as to unload directly from it. This is a matter of much convenience to the workmen. The width of the cellar floor will permit of a row of barrels being placed in front of the bins and yet admit the vehicle. "A series of observations on the range of temperature in the cellar was made during November, December, January and part of February, and the results are summarized below. The ventilator and the supply air flue were all left open from November 1 to December 10. The outside air temperature registered 28° on the morning of the first of November, and the cellar showed a temperature of 46° Fahr, As the month progressed, a period

*

*

*

*

*

*

*

*

*

Virginia Storage House.

457 of warm weather set in, without a drop to freezing from the 7th until the 24th of the month. During this time the temperature often registered above 60° in the shade, with maximum readings considerably higher. The cellar temperature varied just 12° for the entire month, reaching 58° on two occasions, but closing the month at 46°, with outside temperature at 34°. During December the cellar temperature was reduced quite steadily from 45° to 38°, the daily variations being at most 2°. Outside temperature varied considerably, but the range was between 15° and 46°. Quite a number of observations was made on the working of the supply flue and the ventilators. The tests made showed that air passed through the one hundred and fifty feet supply flue in thirty to forty seconds, and the ventilators could be depended upon at all times to keep up a movement of air in the cellar so as to draw a fresh supply. In fact, during the coldest weather we frequently closed the ventilators to prevent the too rapid lowering of the temperature in the cellar. The tests showed that this short flue could not be depended upon to raise the air to a proper temperature when the mercury outside was at 15° or lower. observations showed that the air was, under these circumstances, raised about 20°, varying, of course, with conditions. During January further experiments showed that we could quite easily reduce the cellar temperature to 35° when the outside air registered 15° to 20°. However, the building proves to be lacking in two essentials to hold the cellar temperature stable; viz., it is not deep enough in the earth, and the floor between it and the tool-room above is not properly laid. This floor is made double, of half-inch stuff, while we now see that the cellar-room should also be ceiled in the best possible manner, to prevent interference by outside changes of temperature. The total range in the cellar during January was 35° to 42°. This result was, however, secured by carefully watching the conditions, so as to admit cold air during the night or periods of low temperature and then closing the flues when the outside temperature would act injuriously upon the temperature of the cellar. With the temperature fluctuations which prevail in this region, much attention is necessary to properly control the conditions in the cellar,”

Our

Alcoholic vapor as a fruit-preservative.-In connection with cold storage, it may be possible to experiment with the "new process for keeping fruit fresh,' which was published in 1895 by the Department of State (Bureau of Statistics, by Henry P. DuBellet, Consul at Rheims, France), and distributed by the Division of Pomology of the Department of Agriculture:

"The great difficulty experienced in preserving fruits in their natural state is such that the dealers who make the attempt of furnishing them out of season are compelled, on account of the heavy losses they sustain, to sell their goods at prices which render it impossible for the great majority of families to place fresh fruits on their tables during the winter months. The high prices which fresh fruits command on the city markets increase day by day from the time they are gathered, and have induced orchard and vine owners to run many risks in order to keep their fruits as long as possible before offering them for sale. And it is not surprising that no pains, efforts, and sacrifices are spared to reach the coveted result, when it is known that during these last years, first-class grapes sold from 2 to 4 francs (38.6 to 77.2 cents) per kilogram (2.2 pounds) from September to November, that they were worth 8 francs ($1.53.4) and as much as 12 francs ($2.31.6) in February and March, and 14 francs ($2.70) in April and May.

"These exorbitant prices show plainly how imperfect are yet the means employed for keeping fruits fresh, how few must be the successful efforts, and how many the difficulties encountered. The solution of this question is, therefore, of great interest to all, and the result of the experiments made in the latter part of the year 1894 and lately reported to the Horticultural Society of Soissons by Mr. A. Petit, chief of the laboratory of horticultural researches at the National Horticultural School of Versailles, deserves the attention and consideration of fruit-growers throughout the United States.