-- aas i .£47 isff (pi. d.%. R* -*■» >»£^C /fct. rrt-J &*-*** REPORT OF A COMMITTEE OF THE AMERICAN ACADEMY OF ARTS AND SCIENCES ON VENTILATORS AND CHIMNEY-TOPS. I March, 1848 . CAMBRIDGE : METCALF AND COMPANY, PRINTERS TO THE UNIVERSITY. 1848 . * V * ) ■ \ ♦ • x -A «• jr' • COMMITTEE, Prof. BENJAMIN PEIRCE, “ JOSEPH LOVERING, “ EBEN N. HORSFORD, Dr. MORRILL WYMAN. % IWBGSTTYCENTcR UBHAHY R B r 0 R T. Dr. M. Wyman, from the Committee, appointed at the Oc¬ tober meeting, to make experiments for testing the value of the principal kinds of ventilating apparatus now in use, made a report, of which the following is an abstract. “ The apparatus used in most of the following experiments consists, 1st, of a machine for producing and maintaining a constant and equable blast of air; 2d, of an arrangement for measuring the velocity of the current produced by this blast. “ The air is put in motion by means of a revolving fan of four blades or vanes, each 21 inches long by 10 inches wide, placed upon the ex¬ tremities of radii 13 inches in length. These blades revolve within a cylindrical case, nearly concentric with the axis of the blades, to which the air gains admission by two circular openings 13 inches in diameter, one in either end of the case. From one side of this case, the air, put in motion by the blades, enters a trunk 3 feet in length, and at its commencement 21 inches wide by 18 inches deep, which is gradually contracted until, at its farther extremity, its cross section becomes a square of 100 inches area. To the mouth of this trunk another is fitted, also 10 inches by the side and 3 feet in length. This last was added to avoid any interfering or unequal currents which might be produced by the converging sides of the first. Upon the axis of the blades is fixed a pinion of sixteen leaves, which engages a wheel of eighty teeth, driven by a handle ; consequently the blades revolve with five times the velocity of the handle, or 300 times per minute when the handle makes one revolution per second. This is the ve¬ locity always used in the following experiments, unless otherwise stated. “ To measure the velocity of the blast, a toy marble, .62 inch in diameter, is suspended by a silken thread, to which it is fastened by a little sealing-wax. This thread is 3 feet in length, and the point of 4 (30S) PROCEEDINGS OF THE AMERICAN ACADEMY suspension, over the mouth of the trunk, is such that the marble hangs as nearly as possible in its centre. The handle is made to revolve ac¬ curately once a second, and the deflection of the marble from the point of rest, under the influence of the blast thus produced, observed. The marble is then protected from the blast, and the effect of the blast upon the thread alone observed and deducted from the first result. To as¬ certain the value of this deflection, the following method is adopted. Into a large cylindrical vessel, filled with water, a pipe, an inch in di¬ ameter and bent into the form of an inverted syphon, is so placed, that, while one of its branches rises in the centre of the vessel, an inch above the surface of the water, the other branch rises along the side of the vessel, over which it is bent nearly horizontally. Another and similar vessel 15.5 inches in diameter at the top, 14 inches at the bot¬ tom, and 8.25 inches in depth, is inverted upon the surface of the water in the first. By pressing down this second vessel the contained air is made to issue from the open extremity of the pipe ; and as the areas of the vessel and pipe are both known, we have but to note the time required to empty the second vessel to learn the velocity of the escaping air. The marble is now suspended by the same thread; the * point of suspension being so situated that the marble falls against the mouth of the pipe, and would, if allowed to move freely, hang as far within it as the marble, deducting the effect upon the thread, was de¬ flected by the blast. The second vessel is now depressed with such velocity that the marble is just made to swing clear of the mouth of the pipe, by which its deflection becomes precisely that produced by the blast which is to be measured. “In the case under consideration, the deflection of the thread and marble together was 2.5 inches ; that dependent upon the thread alone, .95 inch. The time occupied in depressing the vessel until it rested upon the top of the inverted syphon, in several successive experiments, was 12.25 seconds. The contained air was compressed .25 inch to produce this velocity, and, as the pipe rose 1 inch above the surface of the water, 1.25 inches were deducted from the depth of the vessel, leav¬ ing an available depth of 7 inches. The mean diameter, that at the top being 15.5 inches, and at the bottom 14 inches, is 14.75 inches. As the areas of circles are to each other as the squares of their diam¬ eters, we have these areas in the proportion of 217.56 to 1. This number multiplied by the depth in inches, 7, gives the whole expendi¬ ture in 12.25 seconds, the time required to empty the vessel; from which OF ARTS AND SCIENCES. ( 309 ) 5 we obtain a velocity of 124.32 inches, or 10.36 feet, per second,— 7.06 miles per hour. This, therefore, may be assumed as a near ap¬ proximation to the velocity of the blast, when not otherwise mentioned. “ The velocity of the induced current being the true measure of the practical value of different forms of ventilating apparatus, it becomes necessary to ascertain this value as accurately as possible. The in¬ convenience attending measurements in which time is involved as one of the elements, and also, probably, the difficulty of determining the instant when a current has passed through a certain space, have led to the adoption of other means, by which the velocity of the current is not directly measured, but inferred. The mode which has been re¬ peatedly adopted, of measuring the efficiency of a ventilator by its power of sustaining a weighted flap or valve, or a head of water, or by some other statical effect, is decidedly objectionable. Such a meas¬ ure gives the correct value of the initial force or tendency to establish a current in a chimney in which there is no actual movement; but it does not indicate the velocity of the current which will be the final result of the action of the ventilator, nor is it any measure of this final velocity when ventilators of different construction are compared together. Mechanics and engineers are familiar with the difference between the statical and dynamical effects of a force. They are aware that the former may be greatly increased by the mechanical powers, so that, through the medium of a pulley or a lever, a single pound may be made to sustain and raise a hundred times its own weight. But the dynamical effect is not correspondingly increased, for in order to raise one hundred pounds through the height of a foot, the one pound must in all cases fall one hundred feet; so that the loss of height precisely balances the gain in weight. In the same way, the dynamical effect of different springs is not to be measured by their strength alone ; it is not simply dependent upon the amount of weight which they will sus¬ tain, but equally upon their length, or rather upon the distance through which they move in restoring themselves to equilibrium. The archer’s bow is a good instance of this assertion, which any one can try for himself, and he will find, that, with a given exertion of strength, he is able to throw the arrow farthest and highest with that long bow of which he can draw the string to his full arm’s length, and not with the strong bow which he can hardly move. But an example more nearly allied to the case under consideration is derived from the air-pump, in which the dynamical value of any amount of exhaustion is equal to 6 ( 310 ) PROCEEDINGS OF THE AMERICAN ACADEMY the power required to produce it, and is, therefore, proportioned to the magnitude of the receiver when other circumstances are the same; whereas its statical power or its power to sustain a head of water is wholly independent of the magnitude of the receiver, and proportioned solely to the tension of the air within it. In all these cases, there is a striking difference between the operations of using the statical and dynamical effects, which deserves the most careful consideration, be¬ cause it is essential and characteristic. The statical effect may be used for any length of time without being impaired, and the reason is obvious ; it manifests itself in a state of rest, when there is no change of condition. The dynamical, on the contrary, can be used once and but once. The one pound can balance the hundred pounds as long as the materials of the pulley and lever will endure ; a compressed spring may sustain its weight, or the expanded air its head of water, as long as we choose, without any diminution of effect. But when work is to be done, a change to be effected, a weight to be raised, a velocity to be produced, the result can only be obtained by a corresponding change in the opposite direction, an undoing of work, a falling of a weight, a consumption of power once and for ever. In the present case, in which the object is to obstruct or divert the motion of the wind in such a way that part of its velocity may be communicated to the air in the chimney, and thus produce a current, the amount of this communication and trans¬ fer of velocity cannot be measured when it does not take place, — when, on the contrary, the mouth of the chimney is entirely stopped up, so that it is impossible to produce any current within it. It would be just as proper to weigh a water-wheel by the weight which will just reduce it to a state of rest, instead of that smaller weight which reduces it to its usual working velocity, and which is universally adopted by experi¬ enced engineers as the correct measure of the power of the wheel. It should also be borne in mind, that there are resistances offered to air in motion by the tube through which it passes. These resistances are not constant; they increase as the perimeter and length of the tube directly, and also as the square of the velocity ; these, it is obvious, cannot be measured where they do not exist. “ The plan, therefore, which has been adopted in these experi¬ ments, is to measure directly the velocity of the current produced, and it will not be surprising, after what has preceded, if some striking dif¬ ferences should be observed between the results thus obtained and those derived from any statical measure. OF ARTS AND SCIENCES. (311) 7 “ To measure the current, a leaden pipe (the material most readily at hand), 1.25 inches in diameter and 53 feet in length, is placed near and a feW inches below the mouth of the blowing-machine. This pipe is coiled, as it leaves the manufactory, into a circle of about 2.5 feet in diameter, of which it makes eight turns. In the mouth of the trunk, before described as attached to the blowing-machine, is a tube of tinned iron, of the same diameter as the pipe, and bent at a right angle ; the upright branch, about six inches long, reaching to the mid¬ dle of the mouth, while the horizontal portion, about five inches in length, reaches within 2.5 inches of the end of the leaden pipe. Each ventilator, when examined and tested, is placed upon the upright por¬ tion of this tube. For this purpose the ventilator has through it, or attached to its side, a corresponding tube of the same diameter. The connection between these two tubes is completed by a glass tube 4 inches long and 2 inches in diameter, and the fitting made close by means of cotton-wool fastened loosely around the extremities of the two metallic pipes. In this compound pipe the current is induced, and its velocity noted. To effect this last object, advantage is taken of the well-known action of iodine upon starch.* Iodide of potassium is dis¬ solved in a strong solution of starch in hot water, in the proportion of three grains or more of the iodide to an ounce of the solution. A piece of paper wetted, or rather smeared, with the prepared starch is suspended within the glass tube, which can be readily removed for this purpose, by means of a wire hook attached to the metallic pipe. A current is now induced by the action of the blast upon the ventilator, and chlorine gas allowed to enter the opposite end of the pipe, which is kept carefully removed from the influence of the blast. The chlo¬ rine is carried along with the current until it reaches the starched paper, which it instantly dyes a deep blue; the chlorine, by its superior affinity for the potassium, seizing upon it, and leaving the iodine free to act upon the starch. “ Chlorine is conveniently obtained for this purpose from Labar- raque’s solution of chloride of soda, and its liberation quickened, if need be, by adding a few drops of sulphuric acid. When the vial contain¬ ing the chlorine is closed by the finger, and held a few seconds in the * The action of hydrosulphuric acid upon moist carbonate of the oxide of lead was first suggested for this purpose, but the chlorine and iodide were judged most convenient. 8 ( 312 ) PROCEEDINGS OF THE AMERICAN ACADEMY hand, its warmth expels the gas more freely, and when the finger is removed it escapes in a jet, which makes the experiment more de¬ cisive. “ In making the following experiments three persons were usually employed; one to keep up a uniform blast, counting the revolutions of the handle by a watch; a second to throw the chlorine into the pipe, and also to observe and declare the moment when the blue color appears upon the starched paper ; the third to note upon a watch the interval between these two events. “ Results of Experiments. “ 1. Air in motion communicates motion to those portions of air at rest in its immediate vicinity. To this phenomenon Venturi, who discovered and explained it, has given the name of the lateral com¬ munication of motion in fluids. “2. A jet of air falling upon any surface is never reflected, but spreads itself out, and forms a thin layer in immediate contact with that surface. It may be admitted as a principle, that fluids do not, under any velocity or any angle of incidence, possess the property of reflection, like solids, and it is, doubtless, owing to the absence of this property that they adhere to bodies against which they strike. In vir¬ tue of this adhesion, a jet of fluid striking a sphere perpendicularly to its surface spreads itself uniformly over both the superior and inferior hemispheres ; a similar jet striking a horizontal cylinder perpendicu¬ larly to its surface completely surrounds it, and does not leave it until the two parts of the jet meet on its inferior border and form one com¬ mon sheet. (Savart, Annales de Chimie et de Physique , Tom. LIV.) “ When a jet of water strikes a truncated cone perpendicularly to its axis, and just above its lower base, it spreads out, covering more than half its surface, and, rising upward, leaves its upper base in a continu¬ ous sheet, vertically in a plane nearly coinciding in direction with that of the sides of the cone, and horizontally nearly in the direction of tangents to the surface of the cone, while a small portion only of the fluid forms two small streams, which drop down from those two points of the lower base of the cone which are at right angles with the orig¬ inal direction of the jet. “ When a jet meets a circular plane at its centre and perpendicular¬ ly, it forms a thin continuous sheet over the whole surface. Both the direction and continuity of this sheet are preserved far beyond the OF ARTS AND SCIENCES. (313) 9 borders of the circular plane, where its edge is thin, but it follows more or less the direction of the curve of the edge, if it is thick and rounded.* (Savart, Ann. de Chim. et de Physique, Tom LIV. p. 119.) “ 3. When a jet of air impinges upon a surface of limited extent, the atmospheric pressure upon the opposite side of the surface, in conse¬ quence of the lateral communication of motion, is diminished, and a current will be established through a tube, one of the extremities of which is placed in the point of diminished pressure, and the other be¬ yond the borders of the surface. This is the important principle upon which the efficiency of ventilators and chimney-tops depends ; it is also important in its bearing on the position of the mouths of air-trunks for hot-air furnaces; if the mouth be placed in a point of diminished pressure, on the leeward side of a building, air may pass outward, especially from apartments on the windward side of the house. “ 4. When a current strikes the extremity of a tube perpendicularly to its axis, motion is produced through the tube towards the current; and when a current already exists in the tube, if its velocity is less than that of the impinging current, that velocity will be increased. “ When two currents of air of different velocities are moving in pre¬ cisely the same direction, the influence of the more rapid current in accelerating that which is less rapid is not so great as when the angle of meeting is between 20° and 40°. When two opposite currents of equal diameter and velocity meet, they form a circular sheet, pei'pen- dicular to the axis of the veins, and the resulting phenomena resem¬ ble those arising when a current strikes a circular plane. If the ve- * A simple demonstration of these propositions may be obtained by means of a card and candle. If a blast from the mouth be directed obliquely against a card, the flame of a lighted candle will be drawn towards the card, on whatever side of it the candle is held. Increasing or diminishing the velocity of the blast does not change the direction assumed by the flame, but only the velocity with which it is drawn towards the card. If the blast be directed perpendicularly upon the centre of the card, the flame, when passed around the edge of the card, will be driven outward at all points; and if the candle be held near the blast, and at a little distance from the plane surface, the flame will, in virtue of the lateral communication of motion, be drawn towards the surface, and yet by the current of air close to and parallel with the card it will be prevented from reaching it. A strong flame may thus be made to play, apparently with great force, upon the hand, and yet not burn it. An illus¬ tration of this principle may often be observed in the narrow pathway, so con¬ venient for foot-passengers, found after a snow-storm, on the windward side of a high and close fence. 2 10 ( 314 ) PROCEEDINGS OF THE AMERICAN ACADEMY locities of the currents are unequal, the greater velocity diminishes the less, destroys it, or inverts it, according to the excess of velocity. The knowledge of this fact leads at once to the interposition of a plate, to prevent loss of velocity in interfering currents. “5. A thin plate placed upon the extremity of a tube, at the proper angle, causes the impinging current to assume a certain direction, and to produce a certain velocity in the tube ; a similar plate parallel to and above this plate does not increase that velocity. “ A cone placed upon the extremity of a tube produces similar changes of direction in the impinging current, and similar movements in the tube, but another cone above the first does not increase the velocity of those movements. “ 6. Beyond certain narrow limits, the velocity produced in a tube by the action of a current on its conical extremity is not increased by increasing the height or diameter of that cone. The full effect of a cone.may be obtained when its lower base is not larger than one half, nor less than one third, the diameter of the flue on which it is placed. “ 7. If a flat truncated cone be fitted to the extremity of a tube, and exposed to the impinging current, a velocity may be produced in the tube of 1.71 feet per second ; if a similar but much smaller hollow truncated cone be inverted and closely secured to the mouth of the first, the velocity in the same tube may by this means be increased to 2.21 feet per second. The same increase of velocity will be produced if the internal cylindrical bore of the first cone be made conical, with its larger base upward. By the addition of this secondary cone, or by the modification of the interior of the first cone, the velocity of the current is increased over that produced by the simple cone nearly in the ratio of 10 to 13, and as the effect is as the square of the velocity, its efficiency is increased nearly in the ratio of 10 to 17. This is the best form of the simple fixed cone, and the most efficient fixed, ventilator , which has been examined by the Committee. Venturi has shown, that, when a conical tube is applied to a cylindrical pipe, the larger base of this conical tube being 1.8 the diameter of the pipe, and its height 9 times the diameter of this same pipe, the expenditure will, with water, be greater for the cone than for the cylindrical pipe, in the proportion of 24 to 12.1. “ 8. A hollow truncated cone, with its larger base closed by a fiat plate, inverted and placed above a cone similar to that last described, will increase the velocity of the current in the pipe upon which it is OF ARTS AND SCIENCES. ( 315 ) 11 placed over that produced by a simple cone nearly in the ratio of 10 to 13. This is one of the most efficient fixed ventilators with a cap which have been examined by the Committee. The form described in the preceding paragraph, with Cisalpin’s plate placed at a certain height above it, is to be ranked in efficiency with that last described. “9. The velocity of the current produced in a pipe, the mouth of which is presented fairly to the blast, is nearly constant, whether the mouth be cylindrical, conical, with its larger base towards the blast, or the reverse. The diminished area exposed to the blast, ip the latter case, is counterbalanced by the increased velocity consequent upon diminished atmospheric pressure within the cone. “ 10. A difference of temperature between the impinging blast and the produced current does not, within the limits observed, influence the velocity of the latter. “ Experiments. “ In the experiments, each ventilator, when examined, is placed upon a perpendicular fixed tube of tinned iron, in the centre of the mouth of the trunk of the blowing machine. This and all other tubes, when not otherwise mentioned, are 1.25 inches in diameter. The velocity of the blast is 10.36 feet per second, or, as indicated by the revolutions of the handle of the blowing-machine, one revolution per second. The time required for the chlorine to act upon the starch, from the moment it is introduced into the pipe, is given in seconds ; the velocity of the current is given in feet and decimals. The direction of the blast is indicated by the > . /&7ZZ7Z7 ^ Fig. 1. “ Experiment 1. fixed tube, Perpendicular Time in Seconds. 73.2 Velocity per Second. Feet. 0.728 Fig. 2. “ Exp. 2. Straight tube, cut off obliquely at an angle of 45 c turned from the blast, opening 40.0 1.325 “ Exp. 3. Elbow ; horizontal por¬ tion one inch long, opening turned from the blast, .... 72.0 Fig. 3. 0.736 12 (316) PROCEEDINGS OF THE AMERICAN ACADEMY “ Exp. 4. Elbow ; horizontal portion 4 inches long, Seconds. Second. Feet. opening turned from the blast, .... “ Exp. 5. Same ; horizontal portion making, with 70.0 0.757 the direction of the blast, an angle of 30°, 46.0 1.152 “ Same; angle of direction 45°, 41.0 1.290 “ Same; “ “ 60°, 43.0 1.233 “ Same ; “ Fig. 4. “ 90°, “ Exp. 6. Elbow turned from 64.0 0.828 1V//A the blast, and having around its opening a plane surface 1.5 inches wide, .... 31.0 1.71 “ Exp. 7. A perpendicular plate 2 inches wide and 1.75 inch¬ es in height, fastened to that side of the fixed tube next the blast, . “ Same plate attached to the fixed tube, but with its edges in the same direction with the blast, ....... “ Same plate on the side of the fixed tube, opposite the blast, ..... no effect in Fig. 5. > \ i i 33.0 48.0 180.0 1.61 1.05 “ Exp. 8. A square plate, 2 inches by the side, on the top of the fixed tube on the side next the blast, . “ Same plate making with horizon an angle of 80°, “ “ “ 75°, « « “ 70°, “ “ “ 67°, “ “ “ 45°, u u u 22° “ Exp. 9. Square plate, 2 inches by the side, with vertical edges .5 inch wide, turned from the blast, and making an angle of 45° with its direction ; the whole plate making an angle of 75° with the horizon, “ Same plate, making same angle with the horizon, but with its edges turned in the opposite direction; that is, towards the blast, . . . . . 31.5 28.2 27.3 29.0 28.7 39.0 65.0 31.0 24.6 1.63 1.87 1.94 1.83 1.85 1.36 0.791 1.71 2.15 “ Exp. 10. A plate 1.25 inches wide at the base, 2 OF ARTS AND SCIENCES. ( 317 ) 13 Time in Velocity per Seconds. Second. inches wide at top, and 2 inches high, with its edges turned towards the blast, as in the last experiment, Feet. gave very nearly the same results, .... 24.7 2.14 “ Exp. 11. A plate 2 inches wide at the base, 1.25 wide at the top, and 1.5 inches high; angles of sides with base equal to inclination of the plate with the horizon, 76° ; placed on the top of the fixed tube, on the side next the blast, its base being raised .37 inch above the mouth of the fixed tube, .... 29.5 1.80 “ A similar plate added to the opposite side of the tube, ......... 28.5 1.86 “ Similar plates on three sides; open side from the blast,.33.5 1.58 “ Similar plates on three sides; open side at right angle with direction of the blast, .... 32.2 1.65 “ Similar plates on four sides, .... 35.4 1.494 “ Exp. 12. Pyramid formed by the four plates, as last arranged, with its base so fitted to the top of the fixed tube that no air could enter by its side, . . 35.5 1.49 “ Exp. 13. Two similar plates, those used in the last experiments, one arranged as in Exp. 10, and the other similarly placed, but raised .37 inch above the first, ..29.0 1.83 “ The influence of the inclined plate, used in several of the preceding experiments, would at once suggest the application of a figure of revo- olution, which would have a similar effect upon the blast, that is, would direct it upward, and thus assist the escape of the current from the tube. A cone is evidently one form which would have this effect. Indeed, the conical chimney-top has been long in use, and its principle often reproduced under slight modifications of form. “ The cone was proposed as a proper form for the chimney-top, and an account of its application published, more than seventy years ago, by Count Cisalpin, in a memoir entitled Description d'une Cheminee et ittuve de Nouvelle Invention. The plan contrived by Cisalpin con¬ sisted of truncated cones of plate or sheet iron, of different sizes. ‘ When this apparatus is to be used,’ says he, ‘ fit to your chimney your first size; it is of no consequence whether the chimney be round 14 ( 318 ) PROCEEDINGS OF THE AMERICAN ACADEMY Fig. 6. or square, provided it have no holes in its sides, and is open only at the top ; if this put a stop to the smoking, your object is probably ac¬ complished, the equilibrium between the wind and smoke is destroyed (nevertheless assure yourself of this by many experiments, made at different times), and then you have nothing further to do than to attach to three sides of the cone three rods of iron, four, five, or six inches long, on which place horizontally a round plate, having a diameter a little larger than that of the cone, to prevent the rain from entering the chimney.’ The adjoining figure is an elevation from the perspective view given in the memoir. “ In 1788, De Lyle de Saint-Martin, a lieutenant in the French navy, again called attention to this form of chimney-top, in a memoir, giving a full description, with drawings, of its construction, and the results of his experiments. The cap surmounting the cone, instead of Fig. 7. being flat as in Cisalpin’s, was also a trun¬ cated cone, but differing in its proportions from that forming the chimney-top. This arrangement, which is here figured from Saint Martin’s memoir, was examined and approved by the French Academy of Sciences, and published in its Transactions. “ Mr. Tredgold, in his treatise on Warming and Ventilating Build¬ ings, published in 1824, and still a standard work, refers to the conical top as one which may often be employed with advantage, when formed in the manner described in fig. 8; and re¬ marks,— ‘ The upper cap prevents down blasts of air, but in a steady horizontal wind the lower cone alone would be sufficient.’ Its mode of action is described and il¬ lustrated by figures, from one of which the annexed cut is copied. For its origin Mr. Tredgold refers to the me¬ moir of De Lyle de Saint-Martin. It will be noticed that the conical cap has, in the last figure, assumed the spherical form. “ The annexed cut shows the same truncated cone, which has, during the past year, been introduced as quite a novelty, the inventor having gone back to first princi¬ ples, and again mounted the flat top. Fig. 8. Fig. 9. 0 'Mfk in OF ARTS AND SCIENCES. ( 319 ) 15 “ It is quite probable, that the conical and pyramidal earthen and brick chimney-tops now and for many years so generally used are modifications of those introduced or recommended by Cisalpin, Saint- Martin, and Tredgold. Time in. Velocity per Seconds. Second. “Exp. 14. A truncated cone, diameter of upper surface 1.25 inches; diameter of lower surface 4.3 inches; height 1.3 inches ; lower surface upon fixed F ee t. tube; upper surface in centre of trunk, . . . 31.0 1.71 “ Exp. 15. Same cone divided into three cones of equal height by planes parallel to the two surfaces ; two smaller cones, ...... 31.0 1.71 “ Smallest cone, ...... 31.5 1.6S “ Exp. 16. Truncated cone, diameter of lower surface 2.1 inches ; height .35 inch ; diameter of flue and upper surface, as usual, 1.25 inches, . . 31.0 “ Inclination of sides to base, in these last cones, the same; 40°. Fig. 10. “ Exp. 17. Cone ; angle of side with base, at bottom 47°, at top 55°, side concave ; diameter of base, 3.7 inches ; height, 1.4 inches, ..... “ Exp. 18. Cone ; angle of side with base, at bottom 44°, at top 64°, side concave; base 4 inches in diameter; height 1.9 inches, “ Exp. 19. Cone, with its cap, made according to the proportions laid down by Saint-Martin (see fig. 7), 35.4 37.6 34.0 “ Exp. 20.* Cone and plate ; inclination of sides to base 45°; diameter of base 2.9 inches ; height .83 inch (see fig. 9), ....... 1.71 1.49 1.41 1.56 1.58 * Dimensions of cone and plate, from which this model was made, as fol¬ low ; _diameter of flue 18 inches ; base of cone 3ft. 6in.; height of cone 12in.; diameter of plate 3ft. 6in.; height of plate above top of cone 9in. ; thickness of plate l^in. 16 (320) PROCEEDINGS OF THE AMERICAN ACADEMY Time in Seconds. “ Exp. 21.* Cone and plate similar to last; base 2.5 inches in diameter; height .62 inch ; inclination of side to base 45° (see fig. 9), ..... 33.0 “ Same cone without plate, ..... 31.0 “ Exp. 22. Saint-Martin’s cone without the cap ; to the upper surface and around the opening a hollow truncated cone is fitted ; height .62 inch ; angle of sides 42° ; larger base of the frustum upward, . 24.0 “Exp. 23. Cone used in Exp. 21, with a hollow truncated cone, .37 inch high, and angle of sides 42°, fitted as in last experiment, ..... 24.5 “ Exp. 24. Cone ; angle of sides with base 48° ; with hollow truncated cone, as in last experiment, . 25.0 “ Exp. 25. Cone ; diameter of lower base 2.5 inches ; diameter of upper base 1.6 inches ; height .55 inch ; internal diameter at lower base 1.25 inches, and diverging to 1.6 inches at upper base, . . 25.5 “ Exp. 26. Cone similar to that used in Exp. 21, with a flat plate, as recommended by Cisalpin (see fig. 6), .7 inch above top of cone; diameter equal to that of base of cone ; on under surface of the plate a hollow cone .37 inch in height, lesser base downwards, 25.0 “ Exp. 27. Square block representing a chimney ; flue inches in diameter ; sides 2 inches ; height 4 inches ; one side towards the blast, . . . 33.5 “ Same, with corner towards the blast, . . . 35.5 “ Same, with a small cone .5 inch high ; angle of side 63° ; side to the blast, ..... 37.5 “ Exp. 28. Same block, with its plane upper sur¬ face inclined towards the blast, at an angle of 3° with the horizon, ........ 37.0 “ Same, at an angle of 10° with horizon, . . 39.0 “ “ “ 20° “ “ . . 87.0 Velocity per Second. Feet. 1.61 1.71 2.21 2.16 2.12 2.08 2.12 1.57 1.49 1.425 1.43 1.36 0.609 * Dimensions of the original of this model: — diameter of flue 8 inches ; diam¬ eter of cone at base 16 inches; height 4 inches; diameter of plate 16 inches, and 4 inches above top of cone. OF ARTS AND SCIENCES. ( 321 ) 17 Time in Velocity per Seconds. Second. “ Exp. 29. Same block ; upper surface horizontal; a square plate, 2 inches by the side, on that side which Feet, is next the blast, ....... 34.0 1.56 “ Exp. 30. Conical tube, open at both extremities ; diameter of larger opening 2 inches ; of lesser ex¬ tremity 1.3 inches; length 4 inches; inclination of sides 5° ; centre of lateral opening 1.6 inches from lesser extremity ; lesser extremity turned towards the blast, ......... 35.0 1.51 “ Same conical tube; lesser opening reduced to .37 inch,. 54.0 0.981 “ Exp. 31. Conical tube, open at both extremities; diameter of larger 3 inches ; of lesser 1.25 inches ; inclination of sides 15° ; length 7 inches ; centre of lateral opening 1.7 inches from lesser end ; lesser end towards the blast, . . . . « 28.4 1.87 “ Exp. 32. Same conical tube, its sides continued until they form a cone, with its apex turned toward the blast, ........ 51.0 1.039 “ Same, with its axis making, horizontally, an an¬ gle of 35° with the direction of the blast, 31.0 1.71 “ Same; axis making an angle of 15° with the blast, 30.0 1.77 tt it tc