I.\-D
July 15, 1930
VXTRIAL AiVD E S G I S E E R I S G CHEIIfIXTRY
which was the experimental limit of error of the apparatus used for checking the resistance. A fairly constant surface of the heating element is essential for accuracy, as a slight change in the surface coefficient of heat transfer will markedly alter the temperature differential between the two elements. For these as well as other minor reasons, it has been desirable to use quartz spirals instead of the bare wires, despite the increased time lag of thermal equilibrium. This time-lag increase, however, is not excessive, since clear fused quartz has a high thermal conductivity. It is rather the increased mass of the element which causes the increased lag. Acknowledgments
The complete instrument, as described. has been installed a t the Harlem Hospital in New York City, as a means for determining the chemical composition of the air in the atmosphere control room used there in the treatment of pneumonia and cardiac patients. The author is indebted to Dr. Grace
237
Lubin and to Dr. Jesse G. M. Bullowa, of the Oxygen Research Fund of XeTv York University, supported by the Committee for the Encouragement of Medical Research, for pointing out the feasibility of the method for these measurements and records. Owing to their need for a recorder, this work was undertaken. Literature Cited (1) Eucken, Physik. Z., 14, 324
(1913).
( 2 ) King, Johns Hopkins Hospital, Bull. 23,277 (1921). (3) Krueger, U. S. P a t e n t 1,698,887 (1929). (4) Krueger, U. S. P a t e n t 1,715,374 (1929). ( 5 ) Ledig and Lyman, J . Clin. Inuesfigalion, 4, 495 (1927). (6) Palmer a n d Weaver, Bur. Standards, Tech. Pafier 249 (1924). (7) Peters, U.S. P a t e n t 1,504,707 (1924). (8) Rabinowich a n d Bazin, J. Can. M e d . Assocn., 16, 638 (1926). (9) Rideal a n d Taylor, “Catalysis in Theory and Practice,” p. 177, Macmillan, 1926. (10) Rosecrans, J . Oplical SOL.A m . , 14, 479 (1927). (11) Schwarze, A n n . P h y s i k , 11, 303 (1903). (12) Weber, Wied. Ann., 10,103, 304, 472 (1880). (13) Winkelmann, Pogg. i l n n . , 153, 481 (1874).
The Baro-Buret 11-Application to Gas Evolution Methods of Analysis‘ Harold Simmons Booth and Newton C. Jones AIORLEY
CHEMICAL LABORATORY, WESTERN R E S E R V EUNIVERSITY, CLEYEL.4RD, OIIIO
The application of the baro-buret to gas evolution both simple and accurate, A S - E 1-0 L U T I O N methods of analysis is described. and may be readily adapted methods of a n a 1ys i s Data are given to show that the baro-buret, when to analysis by gas-evolution may be applied to any careful attention is given to errors and their correcmethods. T h e a c c u r a c y c o m p l e t e r e a c t i o n which tions, gives both rapid and accurate results in quantiof t h e a p p a r a t u s a l l o w s evolves equivalent quantities tative analysis. The errors are evaluated and the corits use in many different of gas at ordinary conditions analyses without modificarections for them are given. of temperature and pressure. tion. A Dartial list of determinations possible by this I n inany Yuch cases the usual method is included. quantitative methods lead to Apparatus e r r o n e o u s results. For inI n the apparatus, as adapted for this purpose (Figure l ) , stance, in the determination of the reducing power of zinc dust for organic reductions, most gravimetric or volumetric methods an accurately calibrated 100-cc. Heinpel gas buret, K , of do not differentiate the metallic zinc from the oxide. By meas- about 15 nim. bore and provided with a two-way stopcock. uring the hydrogen evolved upon treatment of zinc dust with F , acts as the well of a barometer. The upper tube, B, is of acid the zinc present as metal alone is determined. This ap- the same bore as the buret so that the depression of the plication may be extended to the determination of the hy- mercury due to surface tension shall be the same in both drogen equivalent, of other metals above hydrogen in the elec- tubes. I n order to hare a true Torricellian vacuum over the upper level of the mercury in B , the tubes must be dried as tromotive series. These methods provide a n extremely rapid means of described in a previous article ( 3 ) . After filling with 500 cc. analysis, but have not been extensively used on account of of distilled mercury purified b y the method of Booth and t>helarge errors inherent in the usual apparatus. With the Jones ( A ) , the leveling bulb IS lowered until the level in the Lunge nitrometer, for example, leaks in the rubber connections barometer tube is near the bottom, stopcock F is closed are hard to prevent ; measurement of pressure on a separate and the space is evacuated through the stopcock C. Removal level tube and barometer increases enormously the errors in of the last traces of moisture and a considerable portion of pressure reading; accurate and fine adjustment of the mercury the gas adsorbed on the walls of the tubing is thus accomis difficult; and accurate temperature control is practically plished. With the system still evacuated, the leveling bulb impossible. The original Lunge nitrometer (13) has been is slowly raised until the mercury runs oi-er into the overflow modified b y many (10, 1.5, 1.7) in a n effort to decrease the tube, D. Great care must be taken when the mercury is errors or to make the method more accurate for a special entering the capillary tube to raise the reservoir R slowly to determination. However, in general, these modifications are avoid br:aking the glass a t the constriction. The last small unwieldy and are especially designed for a particular analysis. bubble of gas is expanded as it paises out ahead of the merDevelopiiieiit of the Booth “baro-buret” (3) for accurate cury, assuring complete removal of all gas in the tube. The measurement of gas volumes proTides an instrument which is process is repeated until no more bubbles pass oyer the top into the overflow tube D . On lowering the reservoir R again 1 Received October 2 , 1929; revised paper received March 28, 1930. the space over the mercury surface in B becomes a true This paper is part of a thesis submitted by Newton C. Jones in partial fulTorricellian vacuum. The difference between the levels in B filment of t h e requirements for the degree of master of a r t s in chemistry in and in K , measured on an accurate millimeter scale, 1.5 Western Reserve University.
G
-4,lrALYTICAL EDITIOS
238
meters long, set up behind the buret and barometer tubes, now gives directly the pressure on the gas in the buret. Variations in the barometric pressure are taken care of by the mercury in the capillary and overflow tube, D. Capillary E acts as a gas outlet from the buret. One half of a conical joint, G, is blown onto the other capillary froiii the buret at an angle of 40 degrees. The evolution bottle, H , is blown on the other half of the conical joint, the neck inclined a t an angle of 45 degrees. A small vial, J , of 4 to 5 cc. capacity, is made so that it may be slipped down into the bottle H. When the bottle is revolved on the conical joint to the position shown by the dotted lines, liquid in the vial will run out, /E qince the inclination of the bottle is now 5 degrees. The baro-buret is enclosed in a wooden case with a glass front in order to provide a constant temperature air bath. Two therniometers supported alongside the buret tube, K , and barometer tube, B, give the temperature inside the buret case. Gases collecting in the pressure tubing to the reservoir, R, may be removed at any time through the / stopcock on the trap. L. Figure 1-Baro-Buret a s Adapted t o All stopcocks and the Gas-Evolution Methods conical ioint are fitted with clamps ( 1 ) to prevent loosening and coniequent leakage. Rubber grease freed from gases b y melting in a vacuum is used on all stopcocks and on the conical joint. Operation
For analysis, enough sample to evolve approximately 75 cc. of gas measured a t standard conditions is weighed into the evolution bottle, and the small vial containing 3 cc. of the liquid reagent is also slipped into the bottle, care being taken not to spill any of the liquid. Tight connection is made to the apparatus by means of the joint G. Stopcock F is opened to the bottle and the pressure in buret and bottle is adjusted to 760.0 mm. by means of the leveling bulb. The stopcock F is then turned 180 degrees and all air in the buret is forced out through the capillary E. F is again turned 180 degrees and the leveling bulb is lowered, creating a partial vacuum in the buret and bottle. When the bottle is revolved on the conical joint to the position shown b y the dotted lines, the reagent flows out upon the sample and the evolved gas flows over into the buret. Care must always be taken to keep the pressure less than atmospheric, as the rapid evolution of gas may create enough pressure to blow the conical joint apart. When the original temperature of the buret case has been attained b y the gas in the apparatus (l/* hour has been found sufficient) the pressure is adjusted to i60.0 nim. To reduce the gases to the original temperature of the gas in the evolution bottle, the bottle is immersed in a water bath of the same temperature as that in the buret case. This procedure avoids a correction for change in volume, of the unmeasured space over the buret stopcock. Pressure, volume, and temperature
Vol. 2 , s o . 3
are then observed. To assure constant temperature, two consecutive readings fire minutes apart should give the same volume at the same pre--sure. Application t o Gas Analysis
CARBOXDIoxIDE-The the reaction CaC03
application of tlie baro-hret to
+ 2HC1 = CaC1, + Hz0 + CO? 1
ivas first studied. Pure calcite crystals were selected, ground to a suitable fineness. and dried. About 0.3 grain was weighed into the evolution bottle, 3 cc. of constant boiling hydrochloric acid were pipetted into the vial, and the carbon dioxide rva< evolved and measured as has already been deqcribed. The result%are given in Table I. Table I-Application
to Evolution of Carbon Dioxide
EXPT 1 lveight CaC03, gram 0 3533 292 2 T (abs.) p . mm. of Hg 760 0 p (COT. t o 00 C . ) 756 6 Vapor tension of solution 13 0 I' (of C02 only) 743 6 1, cc. 85 4 1'1 (reduced t o N. T. P . ) , cc. 76 9 i'2 (theoretical), cc. 78 6 i'3 (cor. for gas law deviations ) cc. 78 3 98 2 CaC03 (found), per cent CaC03 (cor. for blank). per cc:nt 100 0 1 4 1'3 - 2'1 b l a n k for 23' C . ) , cc.
EXPT 2
EXPT 3
EXPT 4
0 3104 296 4 760 0 756 6 1.3 0 743 6 72 7 65 4 67 1 66 8 97 9 100 0 1 4
0 3114 296 3 760 0 756 6 13 0 743 6 i5 1 67 6 69 3 69 0 98 0 100 0
0 3492 296 4 760 0 7.56 6 13 0 743 6 84 4 76 0 77 7 77 4 98 2 100 0 1 4
1 4
The difference between the theoretical volume fcorrected for gas lam deviation by multiplying the true theoretical value by 0.996) and the actual x-olunie evolved is in each case 1.4 cc., the runs being made at nearly the same temperature and pressure. This constant difference is obviously due to solubility of carbon dioxide in the reaction mixture and may be regarded as the blank for the range of temperatures employed. For higher temperatures the solubility of carbon dioxide will be less and the blank likewise less. The reverse will necessarily hold for lower temperatures. HYDROGEh=-In order to test the method with a more perfect gas, the reaction Mg
+ 2HC1
MgC12
+ HI t
was studied. Hydrogen is as near an ideal gas as can be obtained. The magnesium used was in the form of 99.9 per cent pure wire from sublimed magnesium kindly furnished by the American Magnesium Corporation, Kiagara Falls, S . Y. The results are given in Table 11. Table 11-Application
t o Evolution of Hydrogen EXPT.2 EXPT3
EXPT. 1 Weight Mg, gram 0 0894 297.8 T (abs.) 760.0 p mm. of Hg p'(cor. t o 0 0 756.3 13 0 Vapor tension of solution 743 3 P (of H? only) 92.0 0 , cc. o (reduced t o iY.T. P.), cc. 82 5 2, (theoretical), cc. 82.4 M g (by this method), per cent 99.8 h l g (by ordinary analysis), per cent 99.9
c.)
0.0947 295.3 760.0 756.6 13.0 743.6 96.4 87 2 87.2 99.9 99 9
0.0951 295 2 760.0 736.8 13.0 743 8 96 9 87.6 87 6 99 9 99 9
EXPT 4 0.0909 295.7 760.0 756.6 13.0 743.6 92.7 83.7 83 7 99.9 99.9
It is seen that, in the case of a gas where we may eliminate corrections for deviation from the gas laws and for solubility in the reaction mixture, accurate and comparable results may be obtained without t'he use of a blank. ~ITROGES-~~itht'he idea of testing a reaction evolving a still different gas, the reaction ( 5 ) 2XaN3
+ I2 = 2NaI + 3 N 2 t
using sodium thiosulfate as catalyst, was used. The sodium trinitride was twice recryst'allized from an acid solution and the only impurity seemed to be moisture. The iodine and
I,VDCSTRIAL A S D E,YGI,VEERI.VG CHEMISTRY
July 15, 1930
sodium thiosulfate were Baker's c. results are given in Table 111. Table 111-Application
P.
Analyzed Grade. The
to Evolution of Nitrogen
EXPT. 1 Xveight NaSa, gram 0.1706 T (abs.) 298.0 ?, m m . of H g 760.0 ? (cor. t o 0' C . ) 756.3 Vapor tension ol solution 24.0 P (of Ng only) 732.3 i'. cc. 95.4 is (reduced t o N. T. P,) cc. 84 2 N a N s (by thi4 method), per Sent 95.6 ATaN3 (by gravimetric analysis), per cent 96 3
EXPT. 2
EXPT.3
EXPT.4
0.1410 298.1 760 0 756 3 24.0 732.3 78.8 69.5
95.4
0 1346 298.1 760.0 756.3 24.0 732.3 74 9 66 1 95.1
0.1573 298.2 760.0 766.3 24.0 732 3 87 8 77.5 95 4
95 3
95 3
95.3
The four runs show results strikingly comparable with those obtained b y ordinary gravimetric methods for determining trinitrides (?, 8). Errors a n d Their Correction
I n any determination b y means of this apparatus there are several unavoidable errors. Readings of volume can be made to 0.05 cc. on a n average volume of i 5 cc., of pressure to 0.2 mm., of temperature to 0.1 " C., of weighing, b y the method of swings, to 0.03 mg., giving a maximum probable error of 0.085 per cent. However, the probability is that some of these errors are positive and some negative for any experiment, so that they may be considered to be self-compensating, Errors inherent in the buret were eliminated b y careful calibration with distilled mercury. The corrected pressure reading to compensate for the difference in gravity constant g a t this laboratory and a t sea level may be obtained b y multiplying the observed pressure by 1.000384 ( 9 ) . Seglect of this correction introduces an error of -0.04 per cent, which is negligible. Since the temperature of the mercury in the barometer column is greater than 0" C., the true pressure at 0" C. o n the gas in the buret is less than the observed pressure. All pressure readings must therefore be reduced to 0" C. by multiplying b y a factor Tvhich is the ratio of the density of mercury ( I d ) a t the temperature of the buret case to that a t 0" C. Correction must be made for the gas dissolved in the solution left a t the end of the reaction. Since this solution contains a mixture of substances for which solubility data are not obtainable, accurate values cannot be ascertained. Approximate values can, however, be obtained by running a blank on a sample of known composition and determining the difference between the actual and theoretical yields of gas. This correction need only be applied in the case of gases of appreciable solubility. I n cases where the evolred gas or gases are near or below their critical temperatures, considerable error may be introduced due to deviation from the gas laws. The deviation may be calculated from Berthelot's equation of state:
239
pressure are introduced into the buret through the capillary E. The stopcock F is then turned through 180 degrees to open the evolution chamber to the buret. The volumes a t two different pressures are observed. The process is repeated at least three times for checks. If s cc. is the volume of the space, and vl, v?, pl, p z are the two volumes and pressures, then from the gas laws pv = constant, or p l ( s rl) = pl(s t'?),whence
+
+
s =
a2p2
PI
- alp, - P2
Ahcuracyof this volume determination is obviously increased by using as large a volume as possible in the buret. The volume of the original liquid and, solid or of the liquid afterward, as the case may be, must be subtracted from this volume to give the true volume of the unmeasured space. If the volume of the liquid reaction products is different from the volume of the liquid and solid a t the beginning, a correction for this change in volume must be made on the observed volume. The vapor tensions of the solution before and after the run were determined in order to find whether there was a n y change during the reaction and to find the true vapor tension of the mixture. T o determine the vapor tension of any solution, the apparatus is set as for any determination with the liquid in the evolution bottle. The buret is then used as the bulb of a Toepler pump, gas being pumped from the evolution bottle H out through the capillary E. Five or six puinpings are generally sufficient to give a constant pressure. The vapor tension is then the difference in the levels of the mercury in B and K . Discussion a n d Conclusion
Determinations involving t h e evolution of gases of appreciable solubility, as the determination of carbonates or bicarbonates, may be carried out rapidly and accurately by means of the baro-buret, provided a blank representing the amount of gas still left in the reaction mixture is added to the observed volume. Such a blank may readily be determined by running samples of known composition and, once found, is changed only b y temperature. However, in cases where the evolved gas is practically insoluble, no blank is necessary and percentage composition can be calculated directly from the reduced volume. The results for hydrogen show admirably the real accuracy of the method. I n the case of nitrogen the gas is not wholly insoluble in the reaction mixture, nor is it so perfect a gas as hydrogen, and yet the results obtained are concordant and compare very favorably with those derived b y other methods. The simplicity and speed of the method combined with the greater precision of the baro-buret make this apparatus peculiarly well adapted for the accurate determination of many substances. A partial list of possihle determinations follo\vs: SOMEDETERMINATIONS POSSIBLE BY THIS METHOD
where p , v. and T are the pressure, volume, and absolute temperature measured in any experiment, p , and T , are the critical pressure and temperature of the gas, R is the universal gas constant. and LVis the number of gram-molecules of the gas. For carbon dioxide a calculation ( p = 760 mm., T = 293" K.) shows p~ = S R T '0.995. Should there be a change in pressure or temperature between the beginning and the end of a run, a correction must be applied for change in volume of the unmeasured gas volume in the evolution bottle. I n order to determine this volume, the evolution bottle and vial, perfectly dry, are attached as in any experiment, and 50 cc. of dry air at atmospheric
(1) Estimation of urea ( 2 ) CO(KHZIZ . 5H20 BNaBr . - - 6NaOH 3Br2 = CO2 N2 (2) Estimation of available acetylene and calcium carbide CaC2 2H20 = Ca(OH), C2Hz (3) Determination of ammonia in ammonium salts (17) 2NH3 3NaOBr = 3 H 2 0 3NaBr N2 (4) Determination of nitrous and nitric acids separately (17) HgzSOc 2N0 H2SOa = 2H20 2HN02 2Hg 3Hg2S01 f 2 N 0 2HN03 6Hg 3HnSO4 = 4Hz0 ( 5 ) Determination of nitrous acid in the presence of nitric
+
+
+
+
+ +
+ +
+ + + + + + + +
' (6) Determination of hydrogen peroxide or the standardization of potassium permanganate solutions (12) 2KMn04 5H202 3HzSO4= 2MnS04 &SO4 8Hz0 50%
+
+
+
+
+
ANB L Y TI C AL EDI TI O S
240
Yol. 2 , N o . 3
( 7 ) Determination of cerium in soluble ceric salts (17)
+ + + + + + +
+
+
+
+
+
+
+
+ +
+ +
+
+
+
+
Literature Cited
+
Ce~(S04)3 HZOZ= 2CeS04 H&04 O2 (8) Determination of carbon dioxide, carbonates, or bicarbonates (17) CaC03 2HC1 = CaClz H 2 0 COS (9) Determination of fluorine as silicon tetrafluoride ( I T ) 2CaFz Si02 2H2S04 = 2CaS04 2H20 SiF4 (10) Determination of available chlorine in bleach (6) CaOClz Hz02 = CaC12 H2O 02 (11) Evaluation of pyrolusite (6) iUnO2 HZOZ H z S O ~= hlnSO4 2HzO 0 2 (12) Determination of zinc in zinc dust (16) Zn 2HC1 = ZnCl2 HZ (13) Determination of hydrogen equivalent of metals (14) Determination cf sulfides FeS 2HC1 = FeC12 H2S (15) Determination of alkyl halides (IT) 2CH31 2Zn 2H10 = ZnIz Zn(0H)z CH4 (16) Estimation of the carbonyl group (11)
+
+
(1) Baume and Perrot, J . chim. phys., 11, 57 (1913). (2) Beilstein, "Handbuch der anorganischen Chemie," 4 t h ed., Vol. 111, p. 57. (31 Booth, IND.E s c . C H E MAnal. , Ed., 2, 182 (1930). (41 Booth and Jones, ISD. ENG.CHEM.,19, 104 ( 1 9 2 i ) . ( 5 ) Browne and Hoel. J . A m . Chem. Soc., 44, 210B (1922). (6) Dennis, "Gas Analysis," p. 401. ( 7 ) Dennis, J . Am. Citem. S o t . , 18, 950 (1896). ( 8 ) Dennis and Doan, I b i d . , 18, 971 (1896). (9) Germann and Booth, J . Phys. Chem., 21, 87 (19171. (10) Joyce and La Tourette, J. I s o . EKG.CHEM.,6, 1017 (1913). (11' Kamm, "Qualitative Organic Analysis," 1st ed., p. 172. (121 Landolt-Bdmstein, T a b e l l e n , Vol. I , p. 76 (1923). (13) Lunge, Ber., 11, 434 (1878). (14) Xleyer, Z . axgew. Chem , 7, 231 (1894). (15) Pitman, J . SOC.Chem. I n d . , 19, 982 (1900). (16: Scott, "Standard Methods of Chemical Analysis," 2nd. ed., p . 487. (17) Treadwell-Hall, "Quantitative Analysis," 6th ed., Vol. 11.
An Electrical Conductivity Method for DetermininQ the Moisture Content of wood',' U
Alfred J. S t a m m FOREST PRODCCTS LABORATORY, LfADISON,
n'IS.
A simple, compact, a n d portable a p p a r a t u s for ELECTRICAL reEffect of Moisture Gradients measuring electrical conductivity has been designed sistance method for for determining t h e average moisture content of wood. determining the nioisThe data for Figure 1, a pin type of contact t h e conductivity measureWith which TI ere obtained by uwig ture content of wood with m e n t s can be translated directly i n t o t e r m s of average surface-contact electrodes in the use of surface electrodes moisture content, n o t only when t h e moisture is u n i the ineawrement of the conhas previously been described formly distributed, b u t also when i t is distributed acductance of wood speciniens, (1). I t was shown that the cording t o a n y normal drying gradient. illustrate very well the effect electrical resistance of wood T h e values of moisture content, as determined f r o m of moisture gradients in the increases over a million fold their electrical conductivities, for 160 specimens of 25 piece, and in addition give for a change in nioisture condifferent kinds of wood showed a m e a n deviation f r o m some interesting information tent from the fiber-saturation t h e value determined t h r o u g h t h e loss i n weight upon upon the gradients thempoint, Tyhich corresponds t o oven drying of 0.6 per c e n t absolute moisture content selves.. The straight line is 20 to 35 per cent moisture a n d a m a x i m u m deviation of 1.7 per c e n t absolute for specimens of u n i f o r m content ( 6 ) , to the oven-dry moisture content. T h e effects of surface films of moismoisture content that were condition. Over this range t u r e a n d surface finishes a r e considered. conditioned in constant-hua linear relationship exists bemidity rooms. The curved tween the logarithm of the line is for wet specimensin the process-of drying, a t room temspecific e l e c t r i d resistance and the moisture content. A brief study of the effect of other variable factors upon perature and 30 per cent relative humidity. The deviation the electrical resistance of wood showed that the inherent of the curve from the straight line is due to the presence of is, to variation of the contact resistance between the electrodes and moisture gradients in the drying specirnens-that the wood, variation of species, and changes in ash content the uneven distribution of the moisture in them. From this (within reasonable limits), as well as changes in the density figure it is evident that extreme errors in the moisture content of the wood, are all relatively small in comparison with the determined will result from usiiig the surface-contact type effect of variation of moisture content. On the other hand, of electrodes when appreciable moisture gradients are presboth the presence of surface moisture and a non-uniform dis- ent. If the drjdng of the speciniens had been carried on at tribution of moisture have such a large effect upon the elec- relative humidities other than 30 per cent, the curve ~ ~ o u l d trical resistance that average moisture-content values can- have been shifted up or down and would have become tangent not be calculated from the measurements. The correct trans- to the straight line at, a different value of nioisture content, lation of electrical resistance of mood between surface elec- for the point of tangency is the point of attainment of equilibtrodes into terms of average inoisture content is thus possible rium conditions with the relative humidity preT-ailing during only in the special case where the moisture is uniformly dis- the drying. For this reason the curve is worthless for detertributed. Since this condition very seldom prevails, it is mining average values of moisture content unless the surface highly desirable to use a different form of electrical contact. moisture content or the relative humidity effect'ive during the The type of contact described in the present article largely period of drying is known for the specimen under investigation. obviates this difficulty, thus making the method more gener- General use under such limitations, of course, is out of t h e question. ally applicable. 1 Received February 17, 1930. It' is of interest to determine whether the curve in Figure 1 2 T h e experimental work and the development of the theoretical concan be accounted for theoretically. For this purpose the siderations presented in this paper were completed in July, 1928. I t was normal moisture gradient curves (from the surface to the considered advisable, however, t o mithhold publication until the present center of the stock) for the drying of Sitka spruce at 30 per time.
A
?yT