SD TIMBER PHYSICS. RESUME OF mA^ESTIGATIONS CARRIED OiN IS THE U. S. DIVISION OF FORESTRY, 1889 TO 1898. By FTLTl^EKT TtOTH, ASSISTANT FKOFESSOK, SEW TOJCE .STATi; COhLEUE OF FOBESTJiS, CORNELL VNlVEliSITY. Repiiiitf-a from H. Doc. No. 181, o3th Cong., 3d Sess. WASHINGTON: GOVERNMENT PRINTING OFFICE. 1899. Book // TIMBER PHYSICS RESUME OF INVESTIGATIONS CARRIED ON IN THE U. S. DIVISION OF FORESTRY, 1889 TO 1898. By FILTBERT ROTH, ASSISTAXT rlCOFESHOK, liE^\^ YOllK STATE COLLEUE OF FOIlEHTIiY, CORNELL I.VJVEJiSITr. Reprinted from H. Doc. No. 181, 5.jth Cong., :!(1 Sess. WASHINGTON: GOVERNMENT PRINTING OFFICE. 1899. ^'^ •<^'^ ^^^*^ Tin; WORK IX TIMBI'R PIIVSICS IN THE DIVISION OF FORI'STK\. liv Fii.iiiERr RoTii, J^ale Asaisltuit in Ihr Dii'ision of Vorfstry, HlS'J'OKKUL. As in the case of other materials, exact investigation of tlie properties of wood did not begin until the latter i)art of the eighteenth and the beginning of tlie nineteenth century, when Girard Buffon and Duhaniel du Moncean in France, and Petcu- Barlow, the nestor of engineering in England, laid tlie foundation for tliis inquiry by devisiug suitable methods and working out correct formula' for the computation of the results. As might be expected, the results of this pioneer work, particularly that of the French investigators, were often contradictory, and have to-day little more than historical value. Subsecpiently our knowledge of wood in general, and that of European species in partienlai', was increased by a number of experimenters. Among these, Ohevandier and Wertheim in France, and Niirdlinger in (Germany, stand out conspicuous. Unfortunatelj', their apjiaratus was ci'ude and, in the ease of the French workers, the series was too small to satisfy so complicated a problem, while Nordlinger was obliged to content himself with small and few specimens, owing to a want of i)roper equii)ment. In England considerable money was expended from time to time both by Government and private enterprise, but tlie eagerness of making the matter as practicable as possible led, unfortu- nately, to much testing of large sizes and to the employment of insufficient (because unsystematic) metliods, so that such extreme exi)erinients as those of Fowke and others have really neither furthered science nor heli)ed the practice. In this country the engineering world for a long time relied largely on the results of Furoi)ean testing, and the wood consumers in general depended on a meager accumulation of experience and crude observation concerning most of tlie line array of valuable and abundant kinds of timber offered in our markets. Ignorance and prejudice had their way. Chestnut oak was pronounced uuiit for railway ties, and thus millions of logs were left rotting in the M'oods, though this prejudice had not a single fair trial to supjiort it. "Bled"' lougleaf, or (ieorgia pine, was considered weaker and less durable, millers and dealers were obliged to misrepresent their goods, causing unnecessary loss and litiga- tion, and yet there existed not a single record of a properly conducted experiment to substantiate these views. Gum was of no value. Southern oak was publicly i)roclaimed as unfit for carriage builders, and the ^■iews as to the usefulness of different timbers were almost as numerous as the men expounding them. The engineering world was the first to realize this deficiency, and men like Hatfield, Lanza, Thurston, and others attempted to replace the few anti(iuated and unreliable tables of older text- books by the results jierforined on American woods and with modern aiqjiiances. In addition to these efforts of engineers, Sharpies, under Sargent's direction, in his great work for the Tenth Census of ISSO, subjected samples of all oui' timber trees to mechanical tests, but, since in these tests only a few select pieces represented each species, tlie engineering world never ventured to use the results. As regards the rest of the wood testing in our country, it may be said that it generally possessed two serious defects: (1) the wood was not properly chosen, and (2) the methods of testing were defec^tive, especially witli respect to the various states of seasoning, wood being tested in almost every state from green to dry, without distinction. This is the more 330 AKi 17 1906 TIMBER PHYSICS. 331 remarkable since the important intlnence of inoistnre was recognized and emphasized by both French and (lerman experimenters more than forty years ago.' These facts were fully appreciated by the engineers of our country, as is well shown by the numerous, often emphatic, approvals and recommendations of the timber-physics work undertaken by the Division of Forestry, and by the eagerness with which wood consumers generally seized on all information of this kind as fast as the Division of Forestry could supply the same. SOUTHEUN AND N^ORTHBRN OAK. Though fully planned before, the work in timber physics was really begun in order to decide an important controversy as to the relative value of Southern and Northern grown oak. A representative committee of the Carriage Builders' Association had publicly declared that this important industry could not depend upon the supplies of Southern timber, as the oak grown in the South lacked the necessary qualities demanded in carriage construction. Without experiment this statement could be little better than a guess,^ and was doubly unwarranted, since it condemned an enormous amount of material, and one produced under a great variety of conditions and by at least a dozen ditterent species of trees, involving, therefore, a complexity of problems dithcult enou-h for the careful investigator, and entirely beyond the few unsystematic observations ot the members of a committee on a flying trip through one of the greatest timber regions of the world A number of samples were at once collected (part of them supplied by the carriage builders committee) and the fallacy of the broad statement mentioned was fully demonstrated by a short series of tests and a more extensive study into structure and weight of these materials. From the.se tests it appears that pieces of white oak from Arkansas excelled well-selected pieces from Connecticut both in stiffness and endwise compression (the two most important forms of resistance). Besults oftfsU on NoHhen, and Sonthem white oak made in Wmhinglon University Laboraton,, St. Lewis, Mo., h, I'rof. J. B. Johnson, 1SS9. BeudiDS and cross breaking. Size of test piece 1% by 1| by 24. Compression. Shearing. Test piece. Stiffness. Ultiraate strength. Resistance to shock. Endwise. Transverse. Longitudinal. Where procured. No. Range No. SMoilalus 1 of elas- ticity, ponuda per sqnare inch. Range No. Modulus 3. W. L. 2. b. h' pounds per square inch. Range No. Modulus '"t^^: 1 Range pounds . jj„s per cubic inch, j Modulus ponnda per square inch. Size 1| by 5 inches. Range No. I Modulus pounds ' per square inch. Range No. Modulus pounds per squan^ inch. A. a. I 1 2 9 5 99(1. 000 1, 280, IIOO 3 1 13,700 18.50;) 4 1 1 59 [ 92 7 6,160 5,48U 1 3 3, 400 3 3, 100 1 1,375 1, 500 Average 3 1 1.135,000 1 16, 130 1 76 3 5, 820 1 3,230 j 1 1,468 A.b. II 3 4 6 1,120,000 10 920,000 8 5 12. 3U0 12, 700 6 47 55 11 i) 4,740 4,980 7 4 2,500 2, 800 C 7 i,'325 Average 4 1 1, 020, 000 3 12, 500 1 3 51 5 4, 860 2 2, 650 3 1 1.225 5 6 11 1 850,000 7 1 1, 140, 000 9 11,400 1 2 12,300 7 83 45 8 10 5, 230 4,820 5 8 2,700 2,500 4 1 1,375 2 1 1.540 5 \ 995,000 1 6 11,860 2 64 4 5,025 3 2,600 2 1,4S8 1 B Size: 1| by I j by 18 inches. Size: Ig cube. _. ^ ; 7 8 9 3 1.570,000 8 1, 100, 000 4 1,385,000 6 12,380 2 14, 690 11 11, 240 9 3 11 27 82 19 4 1 5 6,800 7,800 6,800 11 2 9 2,000 3,200 2,300 10 5 11 ' 860 1, 260 825 Average 2 1 1,351,667 2 12, 770 4 43 2 7,133 4 2, 500 1 5 982 10 11 1 j 1,653,000 2 i 1,581,000 4 10 13, 030 11, 590 8 10 30 22 3 2 6,900 1 6 7,700 10 2,000 2,100 i 8 9 1,050 940 1 1.617,001) 4 12, 310 5 26 ] 7,300 j 5 2,350 4 1 995 1 ) For a more complete history see Bulletin 6 of Division of Forestry. 2 See Report of the Division of Forestry, 1890, page 209. fW. = total load al center in i)ound3 W L 3 where ^- =^ length in juclus. .youn-s ,„odulus of elasticity: ^^^r^;^, IJ; Z^^^!^^:^' \^. = l)eii;lit in inches. 332 FOKESTRY INVESTIGATIONS U. S. DEPARTMENT OF AGKICULTUKE. Ih-iicriiition of li'ffotatioD as to statioq, »ite, ami tree ! A. a. I. CouBecti- [ cut upland. ' NunilHT of test piece ] . Kxp«9Uie iu tree Nortli. Heij,'Iit in Ireu i 'Ilutt cut." Position in treo (with relerencc to periphery) Not known. aiae of te.st material ; I I.cn-th 4 Breadth . I)e]>th imeasurt'd acrosH rings). NuniluT of rings Width of ringrt (average) SumnuT wood as a whole Finn hast tissnc Space lost hy lariio ve.-r. Charles Mohr, of Mobile, Ala., an acknowledged authority on the botany of the region, and thus a correct identification was assured. Of each tree entire cross sections as well as the intervening logs were utilized, the former being subjected to examinations into their specific weight (the acknowledged indicator of many valuable technical properties), iuto the amount of moisture contained, into the shrinkage conse(iueiit on drying, and into the struc- tural peculiarities, particularly those structural features which are readily visible and may be utilized in practice for purposes of timber inspection. The logs were sawed and tested according to definite plans in the well-equii)ped test laboratory of the Washington University, St. Louis, Mo., under the direction of Prof J. B. Johnson, a recog- nized authority in engiueering. The first series of test results are embodied iu BuUetin No. 8 of the division, where the strength values for the longleaf pine are fully tabulated and discussed. So eagerly was this bulletin sought by Avood consumers, that an edition of J,000 copies was exhausted in a Short time. Bled and Unbled I'ine. In addition, this series of tests together with an extensive chemical analysis aud physical and structural examination of material from unbled and bled trees, as well as from trees bled and abandoned for five years, re-enforced by an extended study of bled and unbled timber at various points of manufacture, proved conclusively that the discrimination against bled timber was unwarranted, since the bled timber was neither distinct in appearance, behavior, nor strength. To avoid error in so important a matter, and also for a comparison of the three most important turpentine trees — the Cuban aud longleaf with the loblolly pine — the extensi\ e chemical analyses of Dr. M. Gomberg, of the Michigan University, were repeated and extended by Mr. O. Carr, of the Chemical Division of the Department of Agriculture. This series of additiomil chemical RESINOUS CONtKNTS OF PINE. 333 analyses fully substantiated Dr. Gomberg's work, so that it was safe to announce that: (1) Bled timber is as strong as un bled timber; and (2) that it contains the resinous .substances in the same ■miounts and similarly distributed as the wood of unbled timber, so that it seemed to follow as a simple corollary that bled timber is also as durable as unbled, and hence e.iual to the latter in every respect. . . , . ^ , , i ^i The importance of this fact was quite fully realized. Trautwine, in his standard work, the Engineers' Pocketbook, at once placed the fact cm eminent record, and the lumbermen ot the South, as well as all trades Journals, spread the welcome news in every paper and at every opportunity. . ,, . ^, , ,, ,.,,. , The work of Mr. (lombcrg in determining the distribution ot the resin through the diflerei.t parts of the tree is uni-iue in method and classical in its clear scientific procedure and statement. Since the publication in which it first appeared was at once exhausted, it appears proper to repro- duce it in full, leaving out only a few tables, as a part of the most valuable work m timber physics performed under direction of the Division of Forestry: A Chemical Study op the Resinous Contents and their Distribution in Trees of THE LONGLEAF PiNE BEFORE AND AFTER TAPPING FOR TURPENTINE. [By M. (ioMUKRO.] Botanists tell us that resins are produced by the disorganization of cell walls and by the breaking down of starch granules of cells. Chemists believe that resins are oxidation products of volatile oils, the change being expressed by formula as follows: 2C|„H„;+30=C.,„H,„O,+H,O. Whatever view be correct,' one thing is certain, and that is that the formation ol either resins or essential oils requires the presence in the tree of those peculiar conditions which we call vital. The tree must live, must be active, must assimilate carbon dioxide and imbibe moisture, in order that oil of turpentine and rosin be formed. The heart of the tree is the dead part of it. It does not manufacture any turpentine. A part of the oleoresin in it had been formed when the heartwood was yet sapwood, and remained there after the change from sap to heart had taken place. It is also probable that the heart of the tree acts as a .storehouse in which there is deposited a portion of the oleoresin formed m the leaves and sap. • i ^ i i When a tree is tapped for turpentine there are two possible changes that might be supposed to take place: (1) The tree may be considered as jdaced in a, pathological condition, when it will strive to produce a larger amount of oleoresin in order to supply the amount removed. In a few years the ener^ry of the tree will be exhausted and the amount freshly supplied will fall far below the amount of oleoresin drawn oft by the tapping. The tapping will then have to be discontinued. The oleoresin in the heartwood will in this case remain untouched. (2) The oleoresin previously stored away in the heart might, by some unknown means and ways, also be directed toward the wound. 1 • 1 If the first change takes place then, the tapping will have little effect upon the chemical composition of the heartwood. If, however, the second condition prevails during tapping, then of course the heartwood will be seriously atfected for some time afXer tapping, and will contain a much smaller amount of oleoresin than it contained before tapping. Moreover, the tapping may affect not only the amount of oleoresin, hut also the (luality of the new product and the relative distribution of volatile products. „ , , * For this reason the chemical side of the problem has been approached by parallel analyses ot tapped or untapped trees for their relative amounts of turpentine. It was hoped that by a large series of analyses an average might be obtained showing whether tapped and untapped trees difter from each other in that respect. CHEMICAL COMPOSITION OF TURPENTINE. Under the name of turpentine is known an oleoresinous juice produced by all the coniferous trees in greater or less amount. It is found in the wood, bark, leaves, and other parts of the trees. It flows freely as a thick juice from the iiujisions in the bark. It con.sists of resm or resins ' The one view does not e.'ccliide tlie utker. 334 FORESTRY INVESTIGATIONS U. S. DEPARTMENT OF AGRICULTURE. dissolved iu an essential oil; the latter is sejiarated from the former usually by distillation -with steam. There are many varieties ot turi)entine, corre-siioiidinj^ to the different varieties of coniferai, but only three are commercially important, as tliey are the source of the three principal oils of turpentine. (1) Tiie turpentine of PinuK pinmtvr (syn. P. maritima), collected in the southern departments of France around Bordeaux. From it is obtained the French turpentine, which yields 25 per cent of volatile oil. (U) The turpentine from rinus palmtris^ P. twda, P. luterophyUa, collected in the southern sea-borderinf>' States from North Carolina to Texas. From them, principally from the first source, is obtained the English or American oil of turi)entine, which yields 17 per cent of volatile oil. Formerly the 7*. ri to 156° C. (.'!) Some other tcrpeiies of Cii.Hinj snjall portions ; kind uot known. (4) Some polyterpenes of (Cfilla),,; small portion.s; kind not known. (5) Cymene (?) C,,,!!,,; small portion.s, if any; boils at 175^ to 176° C. (fi) Tracrs of formic and acetic acids; produci'd probnbly Tiy atmosjdieric oxidation dnrin<; collection of tnrpontine. ANALYTICAL WOKK. As both till' lo.sjn and the vohvtileoil are easily soluble in chloroform, ether, carbon disnlphide, etc., their separation from wood by any of the above solvents would appear to l)e an easy matter. But an exact (|nantitative determination of the volatile oil presents considerable diliiieulties, and for these reasons: (1) ^Vood (-an not be dried free from moisture without driving off some of the volatile hydrocarbons; (2) the ether extract can not be freed entirely from either without some loss of the volatile oil. If a weighed (juantity of wood shavings is exhausted with either, the i-esidue dried at lOQo 0. and weighed, the total loss thus found will represent: The moisture = H. The rosin = R. The volatile hydrocarbons = T. It is sntticieiit to deteimine two of these factors; the third could then be determined by difference, lint as has been mentioned before, the ether extract can not be obtained in any degree ^SJ I-'h;. 85. — Metlioil lit clicmic;!] :iTi;iIysiH nf tii["l>fiitiiii-. of purity without loss of turpentine. The evaporation of ether in a stream of dry air, as proposed by Dragendoif, fiir the estimation of essential oils in general, does not give satisfactory results with turpentine oil, as Dragendorf himself observed. A weighed quantity of a mixture of rosin and oil, made up in about the same proportions as they exist in crude turjientiiio, was dissolved in a suitable amount of ether. The latter was then evaporated in a current of dry air till the odor of ether was hardly noticeable. The mixture was found to have gained considerably in weight by retaining ether in the thick sirupy oleorosin. It was only by heating at 100° C. for some time that all of the solvent could be driven off, and then the mixture was found to have lost in weight. Repeated trials proved that this method could not be used safely. An attempt was then made to determine the quantities fi'and R, and thus find T by difference A weighed quantity of wood shavings was placed in a small tlask a. The latter was connected on one side with a tray of drying bottles, on the other two OaCl, tubes h and c, similar in size and form. The tiask is immersed in boiling water and a current of dry air is passed through the whole apparatus for one and one-half Ijours. The flask is then cooled and air is i>assed for one and one half hours longei'. It was thought that while b would retain all the moisture and a i)ortion of the volatile com- l)ounds,c would retain about the same amount of the volatile products only. Gain in weight of INVESTIGATIONS INTO RESINS. 837 c subtracted from that of b would then give the moisture H. The sample of wood shavings is then exhausted with ether, the latter evaporated, and the residue heated at about 140'^ to 150° to constant weight; this gives the rosin B. If L be the total loss by extraction with ether, we have L-H+R=T. But it was soon found by experiments upon pure turpentine oil that the two CaClj tubes did not retain an equal amount of volatile oil. The quantity retained depended upon many circum- stances, the chief one being the amount of moisture already present in the CaCL tubes. Even had the tubes retained (pxantities of turi)entiiie oil, this method would still have the objection that one of the constituents was to be determined by ditterence — an objection especially serious when the ingredient to be so determined is small iu comparison with the materials to be weighed. The writer has therefore attempted to make use of a somewhat diflerent principle. A few trials were sufficient to show that the method promised to give satisfactory results. The basis of the method is the same which served for the production of Eussian turpentine oil on a large scale, namely, the distillation of the volatile products from the wood itself, without previously obtaining the turpentine. But instead of condensing the volatile products, their vapors are passed over heated copper oxide, whereby they are burned to water and carbon dioxide. Many trials were made with this iT.ethod upon pure materials and ou samples of resinous wood. As the results were found to be entirely concordant and satisfactory, the method was adopted, and by it were obtained the results presented in this report. DESCRIPTION OF THE METHOD EMPLOYED. A weighed amount of wood shavings is placed iu a straight CaOL tube «. The tube is con- nected on one side by means of a capillary tube with a drier A, which serves for freeing the air from moisture and CO,,. The other end of the tube is connected with an ordinary combustion Method of distiUutiou of turpeutiue. tube h containing granulated CuO. Tiie tube is drawn out at one end as is shown in the figure, and the narrow portion is loosely filled with asbestus wool. The connection is made glass to glass, so that the vapors of distillation do not come in contact with any rubber tubing. The forward end of tlie combustion tube is connected with a CaUl^ tube c, one-half of which is filled with granulated CaCli and the second half with 1^0,,. Then follows a potash bulb d provided with two straight tubes, the first one filled with solid KOH, the second with P.O,-,. The last tube is connected with an aspirator. All the connections having been made air-tight, the connection between the tube a and the drier A is shut off by means of a clamp and the aspirator turned on. When the combustion tube has been heated to dull redness the burner under the air-bath B is lit and the temperature raised to IIO'^-IL'O^ C. The moisture contained in the tube escapes quite rapidly, carrying with it some turpentine oil. The capillary tube at the other end of A practically checks backward diffusion H. Doc. 181 22 338 FOKKSTUY INVESTIGA'I'IONS U. S. DEPARTMENT OF AGRICULTUIiE. or any accuinulation of condensed vapors. In about fifteen minutes all the nioistui'C appears at tbc forward end of the coMibastion tube. Tlio clamp is now opened and a stream of air at the rate of somewhat over one liter an hour is passed through the whole ai)paratus, while the tem- perature of the air bath is raised to 15.">^ to IGO'^ C, and kept at that point for about forty-tive minutes. Toward the end of the oi)eration the temperature is raised to Kio'^ to 170° C. for teu minutes. Then the light under the air bath is turned off and air aspirated for twenty to twenty- live minutes longer. As the air bath is in close contact with the combustion furnace, the whole length of the tube is kept at a temperature above the boiling point of turpentine oil. In this way a coiiiijlete distillation is insured. All the moisture is retained by c, while the CO^ is absorbed in the potash bulb d. The gain of weight in c represents the moisture originally present in the sample of wood plus the water produced in the combustion of the hydrocarbons. The gain in weight of d represents the amount of CO2, derived from the combustion of the volatile products. The tube a is now transferred to an ordinary Soxhlet's extraction apparatus and exhausted ■with ether. The latter is distilled otf, the residue dried for about two hours at lOO^ C, and weighed. This represents the amount of rosin in the sample of wood taken. As has been previously mentioneil, the volatile oil of the oleoresin is not pure australene, Oii,Hii; = (CsH,.)... It j)robably contains some other hydrocarbons, either of the same formula or belonging to the class of polyterpcnes (C5H,,),,. It is clear that whichever they be their percentage comjiosition is alike in all; they all have C= 88.23 per cent, H = 11.77 per cent. Therefore, so far as the combustion of the volatile terpenes is concerned, they can all be represented by the ecpiatiou: (J„,H„, + 140 = 10 00.,= S H.O 130 440 144 In other words, 440 parts of COj are derived from 136 parts of volatile terpenes. 440 rlSO = 1 : X ; X = 0.3091 , i. e., 1 partof CO.. ol)taiii('(l in the combustion represents 0.301) parts of (he volatile hydrocarbons. For every 410 parts of OO3 produced there are 144 parts of H3O formed. 440 :144 = 1 :X ; X = 0.3272, i. o., simultaneously with 1 i)art of CO2 there is produced 0.327 ijarts of HaO. Let the weight of the sample taken = W, Let the weight of CO^ obtained = W, Let the weight of HjO obtained = W", Then — W x 0.309 = T, the amount of volatile hydrocarbons. W X 0.327 = II', the amount of II^O corresponding to the volatile hydrocarbons. W" X — H', = n the amount of moisture in the wood. T II ^ = per cent of T; «t= per cent ot moisture. Thus the moisture, the volatile hydrocarbons, and rosin are obtained directly from the same saimi)le. Where many estimations are to be made, it is of course unnecessary to cool down the combustion tube between successiv'e combustions. The temperature of dutiUatwn. — Some experiments were made to determine at what tempera- ture it is safe to conduct the distillation. Although pure turpentine boils at 15G-lG()o C, yet in open air it (-an be volatilized at a nuich lower temperature, even on the water bath, without any difficulty. Especially is this the case when the vapors are removed as soon as formeil by a stream of air, but it must be remembered that the volatilization of the essential oil directly from the wood might be considerably hindered by the large amount of rosin. A sample of wood distilled by the method outlined above gave the following results at difierent temperatures: 120° 140° 1501^ 160° 170O H,0 — l*er cent. 1.09 11,17 Per cent. 1.18 11.33 7'er cent. 1.3U 11.23 Per cent. 1.20 Per ce7it. 1.32 1 INVESTIGATIONS INTO RESINS. 339 Another sample gave : 160° 180" 1 HjO = Per cent. 4. no 8.79 Per cent. 3.98 The results would indicate that the distillation is practically complete at 160°, and that iJie wood itself does not contribute any CO, by partial decomposio!i at that high temperature; for, should the latter be the case, higher results might bo expected at ISO'^ than at IGOo, and then the sapwood would give much higher numbers for turpentine oil than those actually obtained. Even if this method does not give the absolute amounts of volatile hydrocarbons, yet it certainly gives results very near the truth, and, what is more important, under the same conditions it gives constant results. Therefore, by employing strictly parallel conditions in the analysis of the different samples, results are obtained which can be safely used as indices of comparison of the relative amounts of volatile hydrocarbons in the samples under analysis, MATEKUL FOU ANALYSIS AND METHOD OF DESIGNATION. Material)). — Trees No. 52 and 53, abandoned five years. Trers No. 60 and 61, abandoned one year. Trees No. 1 and 2, not tapped. Trees 54-57, abandoned five year.s. Trees 58-59, abandoned five years. Trees 63-65, abandoned one year. Trees 66-69, abandoned one year. Trees 17-19, not tapped. Generally Disk 11 is 23 feet from ground. Disk 111 is 33 feet from ground. Disk IV is 43 fret from ground. Method of designation. — It was thought best to make a somewhat detailed analysis of a few bled and unbled trees in order to gain an insight into the quantitative distribution of turpentine in the trees. Each disk was divided into pieces of about thirty rings each, the heart and sapwood being kept separate. The number of the disk is designated by a roman iigure, the kind of wood by either .i for sapwood or h for heartwood. The arabic figure which precedes the h or s de.siguates the number of the piece, counting for the sapwood from the bark; for the heartwood, from the line of division between sap and heart. PreiHiration of material. — The first six tables give the results of what might be called "detail" analysis, where each piece of about thirty rings has been analyzed separately. The material for analy- sis was prepared in the following way: A radial section of the disk, about 1 to 2 inches thick, is selected. A piece of 1 inch is cut off transversely, and the strip is then divided into pieces of about thirty rings each. From the freshly cut transverse surface about 15 grams of thin shavings are planed off and placed in a stoppered bottle. The exact amount used for analysis, usually from 3 to 5 grams, is found by weighing the bottle before and after taking out the portion for analysis. The second set of tables, VII to XII, inclusive, give the results of "average" analysis. The material for these analyses was obtained by mixing equal quantities of shavings from the corre- sponding portions of several trees and taking for analysis an average samjjle of tbe mixture. The sapwood furnish one analysis and the heart wood was either analyzed as a whole or divided into portions, 1/t and 2/t, if of considerable thickness. Notes on Tables I to XII. Each table contains a column "calculated for wood free from moisture," giving the per tent of volatile hydro- carbons and rosin obtained by calculation from results actually found. Objections might be raised to this mode flf interjireting the results. It might bo said that the moisture in the wood can not be disregarded, because it is as much an essential proximate constituent of wood as the turi>entine itself is. But since the analyses were not made soon after the trees had been felled, the moisture found in the samples does not represent the original moisture, nor Is 2a IK 3K 47i. Fk;. 87. — Distribution of turpeutino iu trees. (A piece marked 52 HI 2/i moaus tree No. 52, disk III, tlio second piece of the heart t 340 FORESTRY INVESTIGATIONS TJ. S. DEPARTMENT OF AGRICULTURE. -Itelationsbiji of dillerent jtarts (iT sairni (IJak. does it represent eqii.il portions of it in all samplfs. 'I'lir nuniltcrs given in tlie column ''water" are of course suggestive as to the comparative degree of retention of moisture Ijy the diflercut samples, since the latter were all exposed to about the same inlluenccs. But it seemed best to compare the amounts of volatile hydrocarbons and rosin on wood free from that variable constituent; the more so as sometime elapsed between the analysis of the lirst and last samples. The last column in eacli table contains the ratio lietweeu the volatile hydrocarl)ons and rosin. Tliis ratio is multiplied by 1(10, and moans that for cnery 100 parts id' rosin as many parts of the volatile hydrocarbons are found as is Indicated in the column. 'I'liis ratio [ \ is of little value in cases when th(5 amount of turpentine is small, because a very small increase of (he lirst constitnent — an increase within experimental error — will change the quotient considerably. An increase of 0.07 per cent of volatile hydroi'arbons in 60. IV, Is will bring up - from 7.2 to 10. ,\ decrease of 0.07 jier cent in 52, IV, 2» will change - from 2.'). 20 to about Ul. These numbers arc therefore of very little sij;ni1icance when applied to the sapwood of all sanii)les, to entire tree 52, and to some parts of trees (iO and 1, all of which show only small ijortions of turpentine. DISCUSSION OF RESULTS OBTAINED. Kelation of rosin and rolntilc hydrocarbon to moisture. — -The iiinount of moisture retiiiued by ditt'erent samples does not seem to have any direct relation to the amount of oleoresin in these samiiles. Yet in the same tree, or rather in the dift'erent parts of the same disk, there seems to exist something like a relation of the two. This is especially notice- able in tree JSTo. 53. The moisture retained seems to vary in- versely with the amount of oleoresin in the sample. Compare, for example, in "».'> II, Ih, 2h, 3li; in 53 III, l/(, '2h, 'Mi, 4/t; in 53 IV, 2A, 3h, ih. The piece richest in oleoresin is generally the poorest in moisture. But this is by no incan-s a universal rule. Some trees show about the same per cent of moisture in parts widely l 0.75 342 FORESTRY INVESTIGATIONS IT. DEPARTMENT OF AGRICUT.TURE. and 5.67 are the lowest numbers for tlie corresponding constituents, the highest being 3.40 and l(i.2!l, resitectively. Here aguin we have two trees of about the same age, under apparently the same conditions of growth, t;ip])('(l at the .'^ame time and abandoned for the same length of time before felling, and yet dilfering very widely from each other. It is dihicult to conceive why tap- ping shimld have affected the heartwood of these two trees in such a strikingly different manner. If the assum])tion is made that the tapping had drained l)oth trees equally, what explanation can be gi v(Mi Ibr the fact that within one year of abandonment one tree is very rich in turpentine while the other has less than one-fourth as much? Vomparixon of trees ~)S uiid 53 iriih CO and 61. — Compare 53 and Gl. Here we have two trees both very rich in tnrpentine, but while 53 had five years of rest after tapping, (il had only one yeai'. Had the tapping forced the trees to pour out their oleoresin previously stored up in the heart, we should ex]>ect to find in the time of rest the prime factor lor the tree in resuming its natural condition ; but, on the contrary, results of ainilysis show that time of abandonment before felling is of little importance. While we can have a tree very rich in turpentine within five years after tapping, we can also have trees rich and poor even within one year, and trees almost totally deprived of turpentine in the heartwood within five years after tapping. Comparison of 1 with i'. — These two trees had never been tapped, and yet neither is rich in tnrpentine. No. 2 contains about twice as much turpentine as No. 1, the difference becoming smaller as we go up the tree. The highest numbers for 2 are 1.93 and 14.19 for T and E, respec- tively, the lowest 0.8C and 5.89, with an average of about 1 and 7. We can say that there is as much dift'erence between untapped trees as there is between trees that have been tapped. Average analyses. — The average analyses cover IC trees. Thirteen trees furnish four sets of analyses of tapped trees and 3 treses fnrnish one set of untapped. The results obtained are summarized in the following table: Tree No. II. III. liem.arks. T. B. J^-XIOO. T. Jt. I'xioo. 54-57 57-. 19 63-65 66-69 17-19 Per cent. 0.93 .80 .91 .89 .64 Per cent. 5.88 4. "6 5. .32 4.95 2.98 15. 58 19.63 17.18 18 21.37 Per cent. U.5S .82 I'er cent. 3.98 4.29 14.04 19.10 Abandoned 5 veans. 0... Al>andoned 1 year. D.i. Not tapped. .71 3.21 21.76 These results show a pretty constant average number for turpentine in tapped trees. The heartwood of untapped trees is poorer in both volatile oil and rosin than that of tapped trees. And here again it is worthy of notice that time of abandonment is of little importance to tapped trees. The trees that had been abandoned for one year are fully as rich as those that had five years to recover from tapping. Comparison ED AND UNBLED TREES. 343 Tlie explanation offered in the i>recediiig paragraph gains still more probability wb en trees 60 and 01 are compared with each other and also with 52 and 53. The difference between 1 and 2, the results of average analyses — all these are very suggestive of the theory tliat the sap, and not the heart of the tree, supplies the turpentine when the tree is tapped. The fact that the heartwood of trees felled one year after tapping is fully as rich or as poor as that of trees felled five years after tapping, seems to the writer of especial significance, for it shows that the richness of the heart- wood in a tapped tree is independent of time of rest before felling. It is a well-known fact that when a pine tree is cut transversely, liquid turpentine immedi- ately appears on the fresh su:face of the sa]iwood, while the heartwood remains perfectly clear. It would seem as if the turpentine in the sap is far less viscid than that in the heart of a tree. It is probable that the turpentine in the sap is richer in volatile hydrocarbons than that in the heart. (A difference of cell structure and manner of existence of oleoresins may also account for this difference in part. — B. li. F.) It is generally stated that crude turpentine as obtained on a large scale yields from 10 to 25 T per cent of volatile oil. This gives p-= 11.11 to 30, with an average of over 20. This average T is somewhat iiishcr than that for the ,, as found for the turpentine from heartwood of the 21 trees analyzed. Although experimental data are wanting to show conclusively that the difference in the consistency of tlie oleoresin from sapwood and heartwood is due to a difference in the relative amount of volatile oil, yet it is quite probable that this should be the cause. The oleoresin in the heartwood of trees has been produced for the most part when the "heartwood was yet sa])wood. Therefore that part of turpentine which is found in the heartwood is the oldest in age and consequently has been exposed the longest to oxidizing influences of air, which gradually replace the water when the sapwood changes to heartwood. It is the same kind of oxidation and of thickening which takes place when crude turpentine is exposed to the air and sun, or when a T fresh cut is made in the bark of a tree. It is probably for the same reason that ,, becomes smaller as we approach the pith of the tree, because the parts nearest the pith are the oldest. It is difficult to conceive how the thick oleoresin of the heartwood could be made to flow toward the incision when a tree is tapped. It is also difficult to explain by what means tfie tree could change this thick turpentine into a less viscid solution in order that it may flow toward the wound. One would Judge, a priori, from the great difference in the consistency of the turpentine in the heart and sap that only the liquid turpentine will flow when a tree is tapped. Tapping will then have little effect, if any, upon the oleoresin stored up in the heartwood of the tree. A tree whose heartwood is rich in turpentine will remain so after tapping. The writer is not willing to generalize too hastily from so few results and consider them as a solution of the problem. A large number of analyses, devoid of the possibility of chance selection of samples, is necessary before a positive or a negative answer can be given to the question, does the tapping of trees for turpentine affect the subseciueiit chemical composition of the heartwood? But, however few in number the results are, they admit of the following conclusions: (1) Trees that have been tapped can still contain very much turpentine in the heartwood. (2) Trees that have been abandoned for only one year before felling can contain fully as much turpentine iu the heartwood as trees that have been abandoned for five years. (3) Trees that have not been tapped at all do not necessarily contain more turpentine in the heartwood than trees tliat have been tapped. The following diagram serves to show what proportion of each disk was involved in each of the detail analyses, and the results in each case. The right-hand vertical line represents the pith of the tree, the horizontal lines represent the radical extension of each disk, as numbered by romau number, the position of the disk in the tree being maintained as in nature, IV being the top, II the lower, and III the intervening disk. The subdivisions of radii represent the actual divisions of the disk to scale of one-half natural size, the portions to the left of the heavy subdivision line representing sapwood s 1 and s 2; the portions to the right heartwood /(, /*, divided according to the method as indicated above. The fonr columns.of figures over ea('h disk piece represent results pertaining to that piece; they stand in order from the toi) for (1) mimber of rings, (2) volatile 344 FORESTRY INVESTIGATIONS U. S. DEPARTMENT OF AGRICULTURE. T hydrocarbons, (3) rosin, (4) ratio „; (2) and (3) a.s calculated on wood free from moisture. instance, for tree No. 53, disk IV, .s2, we find — For 40 = Nuiu1)er ol" riiiss. 0.40 = Per cent of volatile hydrocarbons. 3. 81=: Per cent of roHiii. T 10.37 = R' 40. 30, 34 33 31. 35. 0.40 tl, 4(1 4 56 4 49 3.86 2.66 3.81 3, '.It; 24 01 22 23 17.74 15.19 Tree 1 10. 37 11.60 1 19 02 1 20 12 1 21.77 1 17.53 No. 53. 40. 37 35. 38. 30. 18. 0.39 42 3.87 3.81 2.10 1.2;» 2.90 3 OS 21.77 20.09 11.97 il.71 1 13.01 1 13 82 1 17.85 1 18.94 1 17.53 1 13, 10 w. 40. 33. 32. 32. 28, U.18 0,19 2.56 4.39 2.22 1,46 0.97 0.96 12.02 24.70 12.30 8.96 1 18.39 1 19.77 1 21.23 _L 22.43 1 18.29 1 16. 33 IV. III. II. 40. 35. 32. 34. 30. 30. 0.26 0.34 0. 15 0.22 0.23 0.26 1.40 1.34 1. 05 1.97 1.72 1.92 Tree No. 52 40. 18,78 1 25.20 1 9.33 11.11 1 13.38 1 13,64 30. 30. 30. 32. 27. 11, 0.25 0,25 0. 15 0.20 0.14 » 0.18 0.18 1.99 1,87 1. 77 1.87 1.86 1.60 1.53 1 12.71 1 13.67 1 8.64 1 10.51 1 7.85 9.05 1 9.26 40. 40. 36 32 35 24 0.30 0.31 30 U.26 17 17 2.19 2.01 2 17 1 83 1 98 1 51 1 13.64 1 13.48 1 14 14 1 14 38 1 8.83 1 II 60 Tree No. 61. 30, 36. 40. 33. 35. 30. 0,22 0.28 3.07 3.49 3.14 1.08 3.01 2.75 13.55 16.29 14.18 8.04 1 7.35 10.20 1 22.05 1 21.42 1 21. 42 1 13.39 35. 35. 36. 33. 30. 35. 0,20 0.26 1.57 2,09 2,92 0.75 3.01 3.11 7.8H 13,57 11. 34 5.67 1 C, ,50 1 8, 36 1 19.85 1 19, 86 1 25. 81 1 13. 28 II. Tree No. 60. 30. 27. 28 30. 40. 0.16 0.24 84 0.41 2.32 2.66 5 35 3.13 1 7.02 1 9.09 1 15 .59 1 12. So 1 30. 34. 30. 36. 36. 20. 0.28 0.35 0,58 0. 40 0.42 0.50 2.65 2,88 3, 60 2.99 2.42 3. 39 1 10.33 1 12.16 1 15.27 I 13,23 1 17.04 1 14.70 30. 35. 37. 33. 35. 27. 0.29 0.33 0.71 0. 51 0. 73 0.47 2.28 2.63 5. 03 2. 71 5. 19 3. 62 1 12.74 1 12. 56 1 14.07 1 18. 62 1 14. 03 1 13. 00 Tree No. 1 . 30. 28. 32. 19 0.22 0.25 1. )7 1 06 1.43 1.57 7.61 6 62 1 15.27 15. 97 1 14. 12 16 04 30. 33. 30. 25. 13. 0,32 0.34 0.94 0.73 0.40 2,25 2, 25 4.90 5,12 3.57 1 14. 49 1 13. 90 1 19. 11 14,21 1 11.20 30. 35 35. 34. 15. 0.20 0,17 0,18 0.66 0..37 1.06 1,32 6,57 3.92 2.23 18.55 1 13.72 1 17.97 1 16.67 1 16. 50 30. 36. 30. 30 0.31 0.34 1.13 87 2.62 2.71 8.10 6 41 1 12.12 12. 36 1 13.98 13 53 30. 36. 33. 28, 17. 0.18 0.24 1.37 0.92 0, ,S6 1.95 2.24 9.14 5,89 7.40 Tree No. 2. 1 8.94 1 10,06 1 14.77 1 15.61 1 11.64 30. 26. 34. 30. 30. 11. 0.20 0.31 1.55 1.93 1.39 1,10 4,29 3.05 10.10 14.19 8.78 8,04 1 4.58 1 10. 00 1 15. 35 1 14.4 1 15, 75 1 12.99 III. II. IV. III. II. FiQ. 90.— Diagram of detail analyses, representing radial dimensions of test pieces in each disk. Scale, onelialf natural size. DISTRIBUTION OF RESINOUS CONTENTS. 345 Table I.— TREE No. 53. Part of disk. Number of rings. Width. Water. Volatile liydro- ciu-boH. Ko.'^in. Calculated ou wood free from moisture. Vol. hydroc. Volatile hydro- carbon. Ko.siu. No. of ili.sk. Rosin, '-'l"" II Ill • IV Is 2s III 2A 3A ih U 2e 1ft 2A 3/1 4A Is 2s lA 2h 3/1 ih 37 40 33 32 32 28 40 37 35 38 30 18 40 30 34 33 31 15 Vm. 3.3 4.0 3.0 2.9 5.0 10.0 2.7 2.6 3.5 4.1 5.5 7.0 4.0 3.0 3.9 3.0 5.8 6.3 Per cent. 10. 51 10. 05 9.1] 8.79 8.47 *11.23 9.08 8.90 7.89 8.04 8.55 8.79 8.96 8.67 8.04 7.93 8.65 9.-55 Per cent. 0.16 0.17 2.32 4.00 2. 03 1.30 0.35 0.38 3.57 3. 50 1.92 1.14 0.30 0.42 4.20 4.13 3.53 2.41 Per cetit. 0.87 0.86 10. 93 17. 83 11.26 7.96 2.69 2.75 20. 05 18.48 10.95 8.86 3.47 3.62 22.08 20.56 16.21 13.74 Per cent. 0.18 0.19 2.56 4. 39 2.22 1.46 0.39 0.42 3.87 3.81 2.10 1. 2;-. 0.40 0.46 4.50 4.49 3.86 2.66 Per cent. 0.97 0.96 12. 02 24. 70 12,30 8.96 2.96 3.02 21.77 20. 09 11.97 9 71 3.81 3.96 24.01 22. 33 17.74 15.19 18.39 19.77 21.23 22.43 18.29 16.33 13.01 13.82 17.85 18.94 17.53 13.10 10.37 11.60 19.02 20.12 21.77 17.53 *53, II, 4ft has been analyzed some three weeks earlier than the remaining parts of this tree, hence a large per cent of moisture. Table II.—TREE No. 52. 1< 40 3.1 9.72 0.27 1.98 0.30 2.19 13.64 2» 40 3.9 9.77 0.28 1.81 0.31 2.01 15.47 II Ih 36 4.6 8.67 0.28 1.98 1.08 0.30 2.17 1.83 14.14 3h 35 6.8 8.80 0.16 1.81 0.17 1.98 8. 83 ih 24 7.4 8.55 0.16 1.38 0.17 1.51 11.60 U 30 3.0 9.12 0.23 1.81 0.25 1.99 12.71 2s 40 3.5 9.00 0.23 1.68 0.25 1.87 13.67 1/i 30 3.4 8.44 0.14 1.62 0.15 1.77 8.64 III 1 2/1 3/1 30 32 3.0 4.8 8.51 8.37 0.18 0. 13 1.71 1.70 0,20 0.14 1.89 1.86 10.51 7.65 ih 27 6.9 9.35 0.14 1.45 0.15 1.60 9.65 ih 11 .5.0 9.21 0. 13 1.39 0.14 1.53 9.26 U 40 3.5 8.88 0.24 1.28 0.26 1.40 18.78 2s 35 3.3 8.49 0.31 1.23 0.34 1.34 25. 20 IV < U 32 3.0 2.8 9.08 0.14 1.50 0. 15 1.65 1.97 9.33 11.11 3/1 30 3.6 8.48 0.21 1.57 0.23 1. 72 13.38 ih 30 6.8 8.10 0.24 1.76 0.26 1.92 13.64 Table III.— TREE No. 61. Is 35 3.0 7.94 0.18 2.77 0.20 3.01 6.50 2» 35 3.0 7.90 0.24 2.87 0.26 3. 11 8. 36 II 1 l/i 36 2.8 7.35 1.45 7.30 1.57 7.88 19. 85 2/j 33 3.2 7.58 2.49 12.54 2.69 13.57 19.86 3/1 30 4.5 7.64 2.70 10.46 2.92 11.34 25.81 4/1 35 9.5 7.10 0.70 5.27 0.75 5.67 13.28 1« 30 3.0 7.65 0.20 2.78 0. 22 3.01 7.35 2» 36 2.7 7.43 0.20 2.55 0.28 2.75 10.20 III I l/i 40 3.1 7.14 2.85 12.58 3.07 13.55 22.65 2/i 33 3.2 7.46 3.23 15.08 3.49 16.29 21.42 ah 35 6.0 7.41 2.91 13.59 3.14 14.18 21.42 ih 30 8.0 7.09 1.00 7.47 1.08 8.04 13.39 Table IV.— TREE No. 60. 1* 30 2.7 9.91 0.26 2.04 0.29 2.26 12.74 2s 35 2.8 9.34 0. 30 2.39 0.33 2.63 12. 56 n l/i 87 3.5 8.72 0.65 4.62 0.71 5,03 14.07 2/. 33 4.5 9.15 0.46 2.47 0.51 2.71 18. 62 3/1 35 4.6 8.01 0.67 4.71 0.73 5.19 14. 02 4/1 27 6.5 8.45 0.43 3.31 0.47 3.62 13,00 Is 30 3.1 8.74 0.25 2.42 0.28 2.65 10.33 2s 34 2.8 8.60 . 0.32 2.63 0.35 2,88 12.16 Ill l/i 30 3.2 8.68 0.,53 3.47 0.5B 3.80 15.27 2/1 36 4.4 9.02 0.36 2.72 0,40 2.99 13.23 3/1 30 4.5 7.73 0.38 2.23 0.42 2.42 17.04 4/1 20 6.0 7.73 0.48 3.13 0.50 3.39 14.70 Is 30 2.6 7.51 0.15 2.15 0.16 2,32 7.02 2s 27 2.6 7.84 0.22 2.45 0.24 2.66 9.09 IV Ih 28 3.7 7.77 0.77 4.94 0,84 5.35 16.59 2/1 36 5.0 8.12 0.37 2.88 0.41 3.13 12.86 3/1 40 8.0 7.92 0.26 2.81 0.28 3.05 9.18 346 FORESTKY INVESTIGATIONS U. S. DEPARTMENT OF AGRICULTURE. Table v.— TREE No. 1. Calculated on woed free from moiature. No. ..f.liak. Part of disk. Number ef ringa. Width. Water. Tolatilo liydro- carlion. Rosin. Volatile hydro- carbon. Rosin. Vol. hydroc'. Rosin. ■ '"" Cm. 7Vr cent. J't-r cent. Per cent. Per cent. J'er cent. l8 30 2.0 8.67 0. 18 0.97 0.20 1.00 18.55 2» 35 3.0 8.77 0.10 1.21 0.17 1.32 13.72 II J \h 3.-. 3.6 8.5U 1.08 6.01 1.18 6.5T 17.97 2/1 34 fi.5 8.39 0.60 3.6re can be used in smaller sizes tlian customary at present, must be welcome as permitting a saviug in fonst resources which may readily be estimated at from eight to ten million dollars annually, due to this information. The pines under consideration, often but imperfectly distinguished by consumers in name of substance, .are: (1) The long-leaf pine { I'iniia jiahiatris), also known as Georgia or yellow jiine, and in England as "pitch pine," and by a number of other names, is to bo found in a belt of 100 to 150 miles in width along the Atlantic and Gulf coasts from North Carolina to Texas, furuishing over i^O per cent of the pine timber cut in the South — the timber par excellence for heavy construction, but also useful for flooring and in other directions where strength .and wearing qualities are required. (2) The Cuban pine (Pinus heteropht/lla), found especiiilly in the southern portions of the long-leaf pine belt, known to woodsmen commonly .as "slash pine," but not distinguished in the lumber market. It is usually mixed in with long leaf, which it closely resembles, although it is wider ringed (coarse grained), and to which it is eijual if not superior in weight and strength. (3) The short-leaf pine (Piiiiis ecldnnta), also known, besides many other u.ames, as yellow pine and as North Carolina pine, but growing through .all the Southern States generally north of the long leaf pine region; much softer and witli much more sapwood than the former two, useful mainly for small dimensions and as linishiug wood, being about 20 per cent weaker than tlio long-leaf pine. (4) The loblolly or old-fielil pine ( I'iuiis tada), of similar although more Soutlieru range than the short leaf, also known .as Virginia jiiue, much used locally and in Wasiiiugtou and Baltimore, destined to find more extensive application. At present largely cut together with short leaf and sold with it as "yellow pine," or North Carolina pine, without distinction, although sometimes far superior, approaching long-leaf pine in strength and general qualities. The names in the market .are often used interchangeably and the materials in tlie y.ard mixed. All four species grow into tall but slender trunks, as a rule not exceeding 30 inches in diameter and 100 leet in height; the bulk of the logs cut at present fall below 20 inches. The sapwood forms in old trees of long leaf (with 2 to 4 inches) about 40 per cent of the total log volume ; in Cuban, short leaf, and loblolly 60 per cent and over. A reliable microscopic distinction of the wood of the four species has not yet been found. As a rule long leaf contains much less sapwood than the other three. The narrow-ringed wood of long leaf (.averaging 20 to 2.5 rings to the inch) usually separates it also from the other three, while the especially broad-ringed Cuban excels usually also by broader summer-wood bands. In the log short leaf and loblolly may usually be recognized as distinguished from the former by the gre.ater projiortion of sapwood and lighter color due to smaller proportion of summer wood. The general appearance of the wood of all four species is, however, quite similar. The annual rings (grain) are sharply detined; the light ycdlowish spring wood and the dark orange-brown summer wood of each ring being strongly contrasted produce a pronouuced piitteru, %vhieh, .although pleasing, especially in the curly forms (which occur occasionally), may become obtrusive when massed. 348 FORESTRY INVESTIGATIONS V. S. DEPARTMENT OF AGRICULTURE. TUo following diagnosis may prove helpful iu the ilistinctiou of tho wooil : Diagnostic features of the wood. Name of species. Long-leaf pine (P'mus Cuban pine {Pinus Short-leaf pine {Pinus Loblolly pine {Pinus ^trta'Linn.). palugtris Miller). heterophylla (VAX) .Sud). evhinata Miller). Specific gravityol'kiIn-/l*o3siblP range . r.0 to . 90 . no to . 90 . 40 to . 80 .40 to. 80 dried wood. (Moslfrequi'iitrangH. . 55 to . C5 . 55 to . 70 . 45 to . 55 . 45 to . 55 Weight, poiiuds per cubic loot, kiln-dried 3C ?>! 30 31 Charatter of grain Been iu cross eeotiun Fine and even; annual Variable ami coarse. Very variable; me- Variable, mostly very rings quite uu i- rings mostly wide; dium, coarse; rings coarse; 3 to 12 rings foriuly iKirrow; on averaging on large wide near heart, iol- to the inch, gener- large logs averag- logs Hi to 20 rings to lowed by zone of ally wider than iu ing generally 20 to the inch. narrow rings; not the short leaf. 'jr> rings to the inch. less than 4 (mostly about 10 to 15) rings to the inch, but often very fine grained. Color, general appearance Even tlark reddish Dark straw color wit h Whitish to reddish or Yellowish to orange yellow to reddish ijrown. tinge ol tlesh color. yellowish brown. brown. Little; rarely overs to 3 inches of radiuft. liroad ; 3 to 6 inches . . Commonly over 4 inches of radius. Very variable, 3 to 6 inches of the radius. Veryabundant; parte often turning into Abundant, sometiuiea yielding more (litch Moderately abundant, least pi"t(!hy; only Abundant; more than short leaf, less than •'light wood;" tlian long leaf; near atuuips, knots. long leaf and Cuban. pitchy throughout. "bleeds" freely, >■ i e 1 d i n g little 'scrape. and liuiba. but does not ' ' bleed " if tapped. The sapling timber of all four species is coarse grained, that of loblolly exceeding the rest in this respect. The grain varies most in the butt, least in the top, is very fine in the outer jtortions of all old trees. Loblolly in the center of the log frequently shows rings over one-half inch wide, .and timber av<'ragiug eight rings to the inch is not rare, while short leaf will average 10 to 1.^ rings to the inch. The greater or less proportion of the sharply defined darli-colored bands of summer wood of the ring furnish the most reliable and re.ady means of determining quality. At present distinction is but rarely made in the species and in their use. All four species are used much alike, although difl'erentiation is very desirable on account of the difi'ereuoe in (juality. Formerly these pines, except for local use, were mostly cut or hewn into timbers, but especially since the use of dry kilns has become general and the simple oil linish has displaced the unsightly painting and "graining" of wood Southern pine is cut into every form and grade of lumber. Nevertheless, a large proportion of the total I'ut is still 1)eing sawed to order iu sizes above ti by (i inches, and lengths above 20 feet for timbers, for which the long leaf and Cuban furnish ideal mati'rial. The resinous condition of these two pines make them also desirable for railway ties of lasting quality. MECHANICAL PROPERTIKS. In general the wood of .all these pines is heavy for pine (31 to 40 pounds per cubic foot, when dry); soft to moderately hard (hard for pine), requiring about 1,000 pounds per sqnare inch to indent one-twentieth inch ; stifi', the modulus of elasticity being from 1,,")00,000 upward; strong, requiring from 7,000 pounds per sipiare inch and upward to break in bending, and over 5,000 pounds in compression when yard-dry. Thi' values given in this circular are averages based on a, Large number of tests, from which only defective pieces are excluded. In all cases where the contrary is not stated the weight of the wood refers to kiln-dried material and the strength of wood containing 15 per cent moisture, which may be conceived as just on the border of air-dried condition. The first table gives fairly well the range of strength of commercial timber. Average strength oj Soutlieni inne. Air-dry material (abuut IJ per cent moisture). Name. Compression streugth. Witli grain. Avei-age of all valid tests. Across Average grain. 'for the weakest, 3 per cent oue-teuth iudeuta- ol' all the tests. lion. Absolute. Rela Absolute, Rela- Bending strength. At rupture modulus f )y' 2 bIC Average of alt valid testa. Average Ifor the weakest modulus I one-tenth | 3 W.i 'of all tlie teats., '2bh^ At elastic Elasticity limit (stiflness) modulus Absolute. Ke^- Absolute. R*- 3 \VP 4 A ill' Relative elastic resili- ence. Tensile Shearing streugth. strength. I Lbs. per «(/. inch. Cuban pine | 7,850 Longleaf pine..| 6, 85C Loblolly pine.. e,5U0 Sborlleaf pine . 5,900 Lbs. per « weight (Transverse strength a) Cuban. 171 6 0.63 39 100 (100) Longleaf. 127 22 0.61 38 97 (91) Loblolly. i:i7 14 0.53 33 84 (84) Shortleaf. 131 10 0.51 32 81 (77) a The values of strength refer to all tests and therefore involve trees of wide range of age and consequently of quality, especially those of longleaf, involve innch wood of old trees, hence tiio rehitiou of weight and strength appears less distinct. From these results, although slightly at variance, we are justified in coududiiig that Cuban and longleaf pine are nearly alike in .strength and weight aud excel loblolly and shortleaf by about 20 per cent. Of these latter, contrary to common belief, the loblolly is thc^ heavier and stronger. The weakest material would differ from the average material iu transverse strength by about 20 per cent and in compression strength by about 30 to 35 per cent, except Cuban pine, for which the difference appears greater in transverse and smaller in compression strength. It must, of course, not be overlooked that these figures are obtained from full-grown trees of the virgin forest, that strength varies with physical conditions of the material and that, therefore, an intelligent inspection of the stick is always necessary before applying the values in practice. They can only represent the average conditions for a large amount of material. DI.STHIISUTION OE WEIGHT AXD STHENGTH THKOl GHOUT THE TREK. In any one tree the wood is lighter and weaker as wo pass from the base to the top. This is true of every tree and of all four S])ecie8. The decrease iu w'eight aud strength is most pronounced in the first 20 feet from the stump aud grows smaller upwaril. (See lig. 91.) This great difference iu weight aud strength betwocu butt aud top finds explanation iu the relative width of the snmmerwood. Since the specific weight of the dark summerwood band in each ring is in thrifty growth from .90 to 1.00, while that of the springwood is only about .40, the relative amount of summerwood furnishes altogether the most delicate and acciiiato measure of these dift'crences of weight as well as strength, and hence is the surest criterion fur ocular inspection of quality, especially since this relation is free from the disturbing inlluoucu of both resin and moisture contents of the wood, so conspicuous iu weight detcrmiuations. The folliiwiug figures show the distribution of the summerwood iu a single tree of longleaf pine, as an example of this relation : At the .stnnip 32 feet from stiunp 87 feet from stump In the 10 rings next to the bark Per cent. 37 25 15 In the 10 rings, Nos. 100 to 110 from bark. 7Vr ccjit. 52 36 37 Average for entire disk. 3'er cent. SO 33 26 Specific weight. 0.73 59 55 350 FORESTRY INVESTIGATIONS U. S. DEPARTMENT OF AGRICULTURE. Weighl and strength of wood at different heights in the tree. Strength of loiifiloaf pine (pouudM por sqtiaro inch). Specific weight. Mejinof all three species. Jtelative weight. Relative strength of lou^leaf pine. Mean of com- pression and bendiug. Bending strength. Couiprea- siou end- wi.se (with grain). Longlcaf. I.obhdI.v. SliorlleaC. 56 150 (over) 22 127 14 113 12 131 48 56 Average age of trees 1 1 1 Number of feet from stump: 1 .751 106 .705 100 .074 06 .624 SU .600 St .500 SO .539 77 .528 75 .629 106 .595 100 .578 07 .534 .90 .508 «6' .491 SS .476 SO .470 79 .614 105 .585 100 .565 97 .523 90 .496 S5 .472 SI .455 7« .454 78 1 ( 1 7,350 urn 7,200 9S 6,800 as 6,500 sn 6,300 SO 6, 150 83 6,050 82 lOG 6 10 12, 100 100 11,650 ■ 100 07 00 SS SI 7.'< 77 im 07 90 SG SI 79 76 20 or, 10, 700 S8 10, 100 SI 9,500 70 9,000 75 8,600 71 30 40 50 60 30 Feet from Stump. Fui. 91. — Variation of weight with height of tree. TIMBER PHYSICS SOUTHERN PINE. 351 60 fT Logs from the top can usiuilly be recognizeil liy the larger percentage of sapwood and the smaller proportion and more regular outlines of the bands of summer wood, which are more or less wavy in the butt logs. The variation of weight is well illustrated in the foregoing table, in which the relative values are indicated id italics. For comparison the figures for strength of loug-leaf pine are added. Both weight and strength vary in the different parts of the same cross section from center to perijjhery, and though the variations appear freciuently irregular in single individuals, a definite law of relation is nevertheless discernible in large averages, and once determined is readily observable in every tree. A separate inquiry, avoiding the many variables which enter in the mechanical tests, permits the following deduc- tions for the wood of these pines, and especially for long leaf, the data referring to weight, but by inference also to strength : 1. The variation is greatest in the butt big (the heaviest part) and least in the tup logs. 2. The variation in weight, hence also in strength, from center to periphery depends on the rate of growth, the heavier, stronger wood being formed during the period of most rapid growth, lighter and weaker wood in old age. 3. Aberrations from the normal growth, due to unusual seasons and other disturbing causes, cloud the uniformity of th(v law of variation, thus occasionally leading to the formation of heavier, broad-ringed wood in old, and lighter, narrow-ringed wood in young trees. 4. Slow-growing trees (with narrow rings) do not make less heavy, nor heavier, wood than thriftily grown trees (with wide rings) of the same age. (See fig. 92.) EFFECT OF AGE. The interior of the butt log, representing the young sap- ling of less than 15 or 20 years of age, and the central portion of all logs containing the pith and 2 to Ti rings adjoining is always light and weak. The heaviest wood in long-leaf and Cuban pine is formed between the ages of 15 and 120 years, with a specific weight of over 0.60 and a maximum of O.GG to ().G8 between the ages of 40 and 60 years. The wood formed at the age of about 100 yc^ars will have a specific weight of 0.62 to 0.63, which is also the average weight for the entire wood of old trees. The wood formed after this age is lighter, but does not fall below 0.50 up to the two hundredth year ; the strength varies in the same ratio. In the shorter-lived loblolly and short leaf the period for the formation of the heaviest wood is between the ages of 15 and 80, the average weight being then over 0.50, with a maximum of 0.57 at the age of 30 to 40. The average weight for old trees (0.51 to 0.52) lies about the seventy-fifth year, the weight then falling oft' to about 0.45 at the age of 140, and continuing to decrease to below 0.38 as the trees grow older. That these statements refer only to the clear portions of each log, and are variably affected at each whorl of knots (every 10 to 30 inches) according to their size, and also by thevarial)le amounts of resin (up to 20 per cent of the dry weight), must be self-evident. Sapwood is not necessarily weaker than heartwood, only usually the sai)wood of the large-sized trees we are now using is represented by the narrow-ringed outer part, which was formed during the old-age period of growth, when naturally lighter and weaker wood is made; but the wood formed during the more thrifty diameter growth of the first eighty or one hundred years — sapwood at the time, changed iuto heartwood later— was, even a.s sai>waod, the heaviest and strongest. Fia. 92.— Schematic section through stem of long loaf pine, showing variation of specific weight, with height, diame- ter, and age, at 20 (aba.), 60 (dcd), 120 (eece), 200 (fStS) years. 352 FORESTRY INVKSTIGATIONS U. .S. DEPARTMENT OK AGRICULTURE. liANCiK <)1' V.\I,l'KS I'OIl WKKiHT AND .-STRENGTH. Although the range of values for the individual tree of any given species varies from butt to top and from center to periphery by 1.5 to 'J5 per cent and occasionally more, the deviation from average values from one individual to another is not usually as great as has been believed ; thus of 56 trees of long-leaf pine, 42 trees varied in their average strength by less than 10 per cent from the average of all 56. The following table of weight (which is a direct and fair indication of strength), representing all the wood of the stem and excluding knots and other defects, gives a more perfect idea of the range of these values: Ilniiije of spcrifio weight wilh aye (kiln-dried wood). [To avoid fractious the values are multiplied by 100.] Number of trees involved . Trees over 200 years old . . . Trees 150-200 y'ear.s olil Trees 100-150 vears old Trees 50-100 years old Trees 25-50 years old Trees uuder 25 years old... Cuban. 24 61 03 61 55 51 Long leaf. Loblolly 96 57 50 60.5 62 61 55 60 50 53 53.4 S3 48 Khort leaf. 51 55 57 53 Though occasionally some very exceptional trees occur, especially in loblolly and short leaf, the range on the whole is generally within remarkably narrow limits, as appears from the following tabic: lianye of apecijic neiglil iu trees of the gome aije {qiproximalely; areragesfor whole trees. (.Specific gravity iiiiiltiplied by 100 to a\'oid fractious.] Name. No. of trees. Age (years). Single trees. Average. r 4 I 5 13 10 12 150-200 50-100 100-150 125-150 100-150 56 68 62 65 ... ... 62.5 60.9 60. .5 52.8 50.8 Lous-leaf piue Loblolly pine Short-leaf pine 59 66 57 62 66 58 59 57 57 66 59 62 57 51 51 53 51 55 5:i 54 55 55 53 45 4.7 53 47 50 51 55 55 53 51 50 53 .. From this table it would a]ipear that single individuals of one species would a)ppr<)ximate single individuals of another species so closely that the weight distinction seems to fail, but iu large numbers— for instance, carloads of material— the averages above given will prevail. INFLUENCE OF LOCALITY. In both the Cuban and long-leaf i)ine the locality Wjhere grown appears to have but little influence on weight or strength, and there is no reason to believe that the long-leaf ])ine from one State is better than that from any other, since such variations as are claimed can be found on any lO-acre lot of timlicr in any State. But with loblolly, and still more with short leaf, this seems not to be the case. Being widely distributed over many localities difterent in soil and climate, the growth of the short-leaf pine seems materially inllucnced by location. The wood from the Southern coast and Gulf region, and even Arkansas, is generally heavier than the wood from localities farther north. Very li"ht and tine-graiued wood is seldom met near the southern limit of the range, while it is almost the rule in Missouri, where forms resembling the Norway pine are by no means rare. The loblolly, occupying both wet and dry soils, varies accordingly. INl'LUENCE OK MOI.STURE. This influence is among the most important; hence all tests have been made with due regard to moisture contents. Seasoned wood is stronger than green and moist wood. The difference between green and seasoned wood may amount; to 50 and even 100 per cent. The influence of seasoning consists iu (1) bringing by means of shrinkage about 10 per cent more fibers into the same s(iuare inch of cross section than are contained in the wet wood; (2) shriukin" the cell wall itself by about 50 per cent of its cross section, and thus hardening it, just as the cow skin becomes thinner and harder by drying. In the following tables and diagram this is I'ully illustrated. The values presented in these tables and diagrams are based on large numbers of tests and are fairly safe for ordinary use. They still require further revrsion, since the relations to density, etc., have had to bo neglected iu this study. TIMBER PHYSICS SOUTHERN PINE. 353 lujlucncc of moisture on stremjih. Avenii;e of ;il! valid tests. Relative vnlnes. Per cent of raoist- ure.a Cu- ban. 8,450 10, 050 11,950 15,300 5,000 6, dOO 7,850 9, 200 "JeSr: Lob- lolly. Short- leaf. Per cent of moist- ure. a Cu- ban. Long- leaS Lob- lolly. Short- leaf. Aver- age. Bending strength Crnaliing endwise f 33 1 20 1 ^^ { 10 1 33 •20 15 10 7, 660 8, 900 10, 900 14, 000 4, 450 5,450 6, 850 9,200 7,370 S, 050 10, 100 12, 400 4,170 5,350 6,500 8, 050 6,900 8,170 9, 230 11,000 4,160 5, 100 5,900 7,000 Bending strengtli Crnsliing endwise — Mean (if both l>eiidiDgaud crushing strength 1 33 1 20 1 15 I 10 ( 33 1 20 15 I lu ( 33 [ 10 100 118 142 181 100 132 157 184 100 125 149 182 100 116 142 182 100 122 154 206 100 119 148 194 100 117 138 168 100 128 156 206 100 122 147 187 100 118 134 160 100 122 142 168 100 120 138 164 100 117 139 173 100 126 152 191 100 122 146 182 a33 per cent green, 20 per cent half dry, 15 per cent yard dry, 10 per cent room dry Variation of coinpiessioii slieii^(h uitli moisture. It will be obseiveil that the strength incrcnses by about 50 per cent in ordinary good yard seasoning, and that it can be increased by about oO pei" cent more by complete seasitning iu kiln or house. Large timbers leiiuiie several years before even the yard-season condition is attained, but L'-inch and lighter material is generally not used with more than 1.") per cent moisture. WKKHir .\Nl> MOI.STUUK. So far the weight of only the kiln-dry wood has been considered. In fresh as well as all yard and air-dried material there is contained a variable amount of water. The amount of water cont.ained in fresh wood of these pines forms more than half the weight of the fresh sapwood, and about one-fifth to oiie-fonith of the heartwood ; in yard-dry wood it falls to about 12 to 1.S per cent, while iu wood kept in well-ventilated and especially in heated rooms it is about 5 to 10 per cent, varying with size of piece, part of tree, species, temper.ature, and humidity of air. Heated to 150° F. ((15° C.) the wood loses all but about IJ to 2 per cent of its moisture, and if the temperature is raised to 175'^ F. there remains less than 1 per cent, the wood dried at 212'^ F. being assumed to be (though it is not really ) perfectly dry. Of course large pieces are in practice never left long enough exposed to become truly kiln-dry, though iu factories this state is often approached. As long as the water in the wood amounts to about .SO per cent or more of the dry weight of the wood there is no shrinkage' (the water coming from the cell lumen) and the density or specific gr.avity changes simply in direct 'In ordinary lumber and all large size material the exterior parts commonly dry so much sooner than the bulk of the stick that checking often occurs, though the moisture per cent of the whole stick is still far above 30. H. Doc. 181 23 354 FORESTIiY 1N\'ESTIGATI0NS U. S. DEPARTMENT OF AGRICULTURE. I>riip()itii)ii to tlio liiss of Wiiti'i'. When tlio inoistiirr i>cr tout falls liilow about 30 the water coiiii'S from ttie itOl w:iil, :iu(l tUr loss of wutc^r ami weight is incoiiipaiiuul by a loss of volunio, so that both factors of the fraction Specilic gravity weight volume ari' adV^ctcd and the change in the specilic gravity no longer is siMii)ly proportional to the loss of water or weight. The loss of weight and volume, however, being uneciual and disproiiortionate, a marked redriction of the sjieeilic gravity takes place, amounting in these pines to about S to 10 per cent of the specific weight of the dry wood. SI11CI>'KA(1K. The behavior of the wood of the southern jjincs in shrinkage docs not difl'er materially. Generally the heavier wood .shrinks the most, and sapwood shrinks about oiu'-fonrth more than heartwood of the same specific weight. Verv lesinons pieces ("light wood") shrink much less than other wood. In keeping with these general facts, the slirinka'^e of the wood of the up])er logs is usually l."i to 20 ])er cent less than that of the butt pieces, and the shrinkage of the heavy heartwood of old trees is greater than that of the lighter peripheral ])arts of the same, w hile the shrinkage of the heavy wood of sa[)lings is greatirst of all. On the whole, the wood of these pines shrinks about 10 per cent in its volume, 3 to 4 per cent along the radins, and (! to 7 per ( ent along the tangent or along the yearly rings. After leaving the kiln the wood at ou<:c begins to absorb moisture and to swell. In an experiment with short pieces of loblolly and shortleaf, representing ordinary llooring or siding sizes, these regained more than half the water and underwent more than half the total swelling during the first 10 days after leaving the kiln (see tig. til). Even in this less than air-dry wood the changes in weight far excel the changes in volume (sum of radial and tan"ential swelling), and therefore the specilic gravity, even at this low per cent of moisture, was decreased by drying and increased by subse(iuent absorption of nu)isture. Innnersion and, still more readily, boiling, cause the wood to return to its original size, but temperatures even above the boiling point do not prevent the wood from "working," or shrinking, and swelling. ^o / B9 jna.) ja.i; •»( n ft ^ e« f^i I- o >C ftSSftffi>CSC5CSXC 111.0 87.6 88.1 104.4 07.0 107.1 102.0 104.S Ul.s e ■!jq3(9M. jiJuha puiii q.iua Siit.viS '9Ioi[ii eqj JO 81 puiJi l(0B9 ■JTI9JJ9J to -K C3 O p4 •pa!!i sjt JO B[ dnoiS qoE» i)u9o J9-5l-5 '. S ^ S =2 . tib >> ib t-. si >) til >i > •" > _" t.' t-j WW o 5«' = ■99J^ JO Jaqnin^ ' © 3 — "^ — O © 3 M r-l CJ ,-1 iQ m tn in iltJioaiJ JO Hiiot)tpnoj |B.x>i •paxoqari ■paxoquj^ 'iD9ioj9q &jB9iC9Ag poxo(i 358 FORESTEY INVESTIGATIONS U. S. DEPARTMENT OP AGRICULTURE. 3 K H o "v. z •^ r^ o ~ O ,J C s h ^.+j w .,. a 2 o ^o M H ss « o S', := « a b^ e "^ •Qioq.vv [I JO H[ oajj q:iB.l 'JlIOD Jnj 104.;! 104.4 110.3 108.7 10(t.7 102.8 10!). 4 •juStOAi tun 1)0 pniT[ t[ob9 3aiAi3 '9Ioi(aC oqi JO K[ pui:^ q.i«a ^aao jaj o ■l>(i!i( KJ! JO Ri dtiojS i[.));^> Jii^n jti.l s »» T* •pupi s)i JO 81 oajj i[ot!o ^na:) joj oooaooooo 11 -- i-« W iH 11 t-l •K an^^^'X M -t- CI t» CO '"' :£ "^ X ©« ifflW — .^ - g&.s "fin J- iC — 1 2 « 4J c ■^ ft -^ S "O CCi CC- I'. 1* i' CC "O ■* .......... 1^ rHi-« «» ft -^ o; in i» « © lo X c ^1 f^ !• '•■^ « « *> «= © " «s t- X C-. -f .-< »M tn ■- ITS • o ^ «» _: X X 1 »SU5 »0 X v« jad osiAvpuo I'ljSiida^e Sinqeiuo »0 ifl .0 >« M r» r» *"" !■• ©9J M.9 I- I. 00 a; i OB , "^1^ S ., .oic/-. QO:DMael^cDCIO^-cS'-lr-«lO-fO'0 9^■*c■ :;^o ©"o x*"*© o c:'^©"cs « « s „ - - «0 WOCl»« © la© ^ X 50 -, ® ® „ m t» ^1 •-» r. aixcJ©-f—xtoo^ooc;(N-HOOoo©or- •i S o g U g -3 g S ^ 1; sS § S ?j I3.S *J 5 atrx I-' 91 ■"■ *• ■iij }im) X'jiAiua oyioails eSBJ^iAy t^ 00 «S t~ if5 CO I~ ^ O !D00.-iMM-*'Oio m m ■ u^ ■•J 'Boqiini lit UdisiiQuiin ttimurxo.icidY ggggooog : ■A y. y. y. -^ \^ >i y. o I f •pa'jBO'; e5[0!'}8 jo .taqinn^ irai-(»ooQoaiT}«co • M OS ScOOOOO ' -00 'S i CO 'co jco • • o£oSS!=^oSo;3o3o3oP _ 1 c 1 a E. a < t- ct 4 1 a & 1 cr. '" 6 c * t 1 c c D t (- H > C <1 F- t C . 6 ti t 1 1 5 = .£ 5 •2 £ < H O o of ft 1 " i IK "H c o < I' c c t- t c J 1- t a a « III m » 50 00 S 00 00 CO 00 OO OO 00 00 -3^ CO CO 2 11 m ■.j^Y a-0 •adJijo jaqranx SSSSSBSS §, •I %i\oiit JO eao[')ipiioo i«noi ■5 TIMBER PHYSICS REAMS AND COLUMNS. 359 SIZE OF TEST MATERIAL. The long- standing idea of engineers and other consumers to have wood tested more nearly in the sizes used in ordinary practice led to the adoption of test sizes, generally varying from 3 by 3 inches to 4 by 4 inches. Besides this, special inquiries with different kinds of timber into the relation of large and small tests were instituted to ascertain the correctness of the general dogma which claimed that tests on small pieces could not be utilized, since such pieces for their very size gave higher values of strength. This investigation involved fall-size columns as well as beams, and was continued throughout the entire period of the timber-physics work. It led to a number of the most interesting and highly valuable results, as will appear from the following statements: Selected teats of columns and comjyreasion pieces from the same trees compared. Ratio Small pieces Large Number of tree. LeDgth. I d (average of whole tree). columns. Relative value. Deflec- tion. Failure. (a) (b) «j) w Feet. Pounds per sq. inch. Pmmdsper sq. inch. Inch. 239 12 U 6,700 6,100 100 91 0.7 Shc-vrod. :;40 ]2 14 7,000 0,900 100 99 0.1 Comitrcssioii. 241 12 15 (i, 900 6,500 100 94 0.7 1)0. :i09 12 12 6,800 6,500 100 96 0.4 Do. 312 ]2 10 6,100 0,300 100 lO.T 0.4 Do. In thfso coliiiiius (nearly ono-tentb of all longloaf pine columns tested) the strength was so nearly the same as that of the short pieces that it appears .as if flexure had hut little to do with the failure, the small dift'erences hcing amply accounted lor hy a larger nuniher of defects in the columns. .Should this prove true in general for wooden columns as ordinarily designed, the problem would become simply a study of the influence of defects and of proper inspection. The nature of the failures would also point in this direction: Of 86 columns 32 failed noruially, i. e., in simple compression ; 22 were crushed near the end ; 14 failed at knots, and 19 by shearing, the rupture usually beginning at or near the ends; a small knot proved sufficient to cause a large column, 20 times as long as its dhameter, to fail at 14 inches from the end. The delleition in the average for all columns (12 to 20 feet long) w.as only .about 1 inch for the maximum load, when, to be sure, destruction h.ad progressed for some time; at the elastic limit the deflection w.as only about one-half as much. These results would seem to warrant the statement that for pine i olumns at least, in which the ratio of height to least diameter does not exceed 1 in 20, none of the accepted column formuhe .are applicable, the ii.ature of the failure being mostly in simple compression, and depending more on specilic defects than on the design of the column. STRENGTH OF LARGE BEAMS AND COLUMNS. Owing to the fact that muih wood testing has been done on small, select, and perfectly seasoned pieces, usually from butt logs, the values thus obtained seemed to dificr very markedly from the results on large tiii-bers usually very imperfectly seasoned, and it was claimed that tests on small sizes always furnished too high values, just as if the ditt'erences were due to sizes alone. While, to be sure, a small piece may be so selected that defects are excluded, the grain straight and in the most favorable position with regard to the load, the assumption of the difterence in strength of small jiieces from that of large-sized sticks has never lieen nuade good experimentally. Since it appears desirable to compare the results from Large beams and columns not only with the .aver.age d.ata obtained from the general test scries on small 4 by 4 m.ateri.al, but also with the average strength of small pieces cut from the s.ame beams and columns, a special inquiry into the h'gitimacy of sucli .a comp.ari.son w.as made. This study involved over 100 separate tests, .and proved the very important fact that uninjured parts of broken be.ams and columns do not sutler in the test. The large-sized beams varied from 4 bv 4 to 8 by 16 inches. Tests of large and small heams — Bcndiny streiujth. Sill.aU beams. general teat series. Largo bciras. Small be.ams cut from large beams. Total. Beams from wbich sm.all beams were cut. Number of tests involved Lon*^leaf 1,986 127 57 236 Lbs. per sq. in. 11, 300 10,000 9,300 Lbs. per sq. in. 11, 500 10, 800 9,200 Lbs. per sq. in. 9, 800 10, 300 8,700 Lbs. per sq. in. 10, 100 10, OOO 8,700 Loblolly Shortleaf From the preceding table it would appear that Large timbers, when symmetrically cut (i. e., with the center of the log as center of the beam), develop as beams pr.ictically the s.ame strength as the aver.age of the small pieces that may be cut from them, and sometimes even higher values ; the explanation being that cut in this manner the extreme fibers which .are tested in a beam come to lie in th.at part of the tret^ which, as a rule, contains the strongest timber. 3(50 FORESTRY INVESTIGATIONS U. S. DEPARTMENT OF AGRICULTURE. Results (lisrordant frniii tlicsi' may ^»^ i-xplained by differences in the degree of seasoninfj of tbe onfer layers and alsd Iiy tlie fact that especially in the northern pineries timlx^rs are often c iit from the top logs, which are weaker and mure defective. 'J'cxI of luriji' Hiid small culntniis — ( 'umpresshin strriiijth. Regular series from same trees as thecolumus. Columns (sim- ple compres- aioD). Small pieces cut from columns. 949 95 97 Lbi, per sq. in. 6,600 6,800 5,900 7,400 Jjhg. per sq. in. .I, 300 4,700 4,100 5,000 Zihs. per tq. in. 7,100 6. 3U(I 6,200 8,700 LobloUv The square columns were mostly 8 by 8 inches, some 10 by 10 inebes, a few of larger and also some of smaller dimen- sions. The ratio of length to width varied from 12 to 27, abont one-half being nuder and the other half over 18 to 1. The compression pieces of the regular series, and those cut from the broken columns, were in general .about 4 by 4 by 6 inches. It will appear from this statement of average results that columns develop only from 62 per cent (in Cuban 1 to 78 per cent (longleaf) of the compression strength id ordinary short pieces. The explanation may be due to several reasons, natural and mechanical. In a column, unlike a beam, all the fibers are under great strain; hence all the defects, which .are by necessity found in every column, inlluenoe the results; the llexure of a column under strain is an element of weakness, to which the short compression pii'ce is nut subjeit. In addition tlie difliculty of determin- ing the average moisture condition of the large timl)er throughout the cross section anortions of the large beam. (See diagram at head of ' The legitimacy of using such material for such purpose has been fully established by a long series of experi- ments. (See Circular 12, Division of Forestry, p. 11.) TIMBER PHYSICS SIZE OP TEST MATERIAL. 361 table.) To avoid any complications due to ditterences or changes in moisture, the tests on large and small beams were performed the same day. Strength of larj/e beams and of umall beamx, and of compression pieces cut from them. [Tlanally 12 small beams cut. from the iiuinjured part of each large beam.] \/ BUTT 6" S' /.Aff6£ B£A/I^ S' 6" TOP 6" 6' 1\ SPI/TS/IIV£0 3526 2728 / 2 3 4 S 6 7 8 5 /O /2 SAWfu.SPi/r TV J 9 20 2122 S32't 2930 3/32 Kiu.l of wood. Number of beam. Strength of large beams. Average strength of small beams. Moisture. Compression, endwise strength. iarge beams. Small beams. Sawed pieces. Split pieces. lihs. per Lbs. per Lbs.%ter Lbs. per so. in. sq. in. Per cent. Per cent. sq. in. sq. in. Oak 2 7,400 5.880 8,560 8,660 69.5 70.3 68.5 69.0 3,960 4,340 4,120 4,700 i 6,570 0. 220 75.3 75.:; 3, 030 3,190 4 8,640 8.800 66.6 67.6 4.0S0 4,460 ,5 8,150 7,710 64.8 65.8 3, 680 3,750 6 7,450 6,910 63.0 66.6 3, 330 3,330 8 6,870 6,890 67.4 70.5 3,470 3,190 Shortleaf pine. 9 10 8,300 7,440 7,950 7, 250 48.1 42.1 57.7 56.3 4,030 3,840 4,160 3,850 11 5,110 6,760 38.9 .33.3 3,870 3, 630 12 7.360 6,930 35.2 33.5 3,890 3,850 13 7,320 7,300 37.4 40.6 4,090 3,800 White pine 14 3,110 3,560 84.9 83.6 2,440 2, .100 15 4.280 4, 340 43.8 41.2 2,710 2,840 16 3,770 4,590 50.7 50.5 2,660 2,760 17 3,460 3, 590 60.0 48.6 2,410 2,570 18 3,990 3,640 42.8 43.0 2,800 2,620 19 4,040 4,400 62.4 60.4 2, 760 2,780 2U 4,410 4,180 53.6 51.8 2,680 2,700 21 4,900 4, 320 50.1 51.0 3,010 2,900 22 3,860 4,320 50.2 60.8 2, 500 2,430 23 4,660 4,890 52.0 58.2 2,850 2,880 24 3,960 4,440 76.3 71.5 2,520 2,710 25 3,920 4,410 53.6 60.5 2,840 2,730 Shortleaf pine 26 4,560 4,390 6, 290 31.2 30.5 3,660 3,850 27 5,610 33.9 36.0 2,830 3,110 28 6,670 6,830 28.6 28.9 3, 540 3,590 29 7.410 7,630 28.6 29.0 4,450 4,250 30 6,600 7,160 28.3 28.9 4,200 4,190 31 5.750 6,000 34.3 35.5 3,630 3,530 32 6,210 7,500 26.4 27.2 3,940 4,050 33 7,450 8,390 29.5 30.1 4,350 4,220 34 7,000 7,800 28.4 29.5 4,070 4,120 35 6,030 «, 740 . 28.8 29.4 3,810 3,640 36 6, 520 6,890 31.6 31.6 4, 320 4,370 37 7,030 7,890 29.2 29.9 4,380 4,920 38 7,710 8,510 26.2 25.4 4,500 4,610 39 8,090 8, 210 32.5 31.9 4, 550 4,670 40 7,680 7.980 31.1 32.3 4, 290 4,380 41 7,330 8, 230 31.7 31.5 4,680 4,820 I.ongleaf pine 42 7,290 8.740 30.9 31.2 4,950 5,120 43 sisso 9!720 28.1 28.9 5,300 5,440 44 8,040 8,870 26.3 26.9 4,730 5,070 45 8,000 H,850 25.8 25.4 5.000 5,050 46 7,620 7,670 32.6 33.9 4,730 4,830 47 6,710 7,610 33.0 33.4 4,200 4,520 4R 8,480 8,300 29.3 29.3 4, 870 4,890 49 5,630 0,250 34.5 33.7 3,600 3,630 White pine. ......... 50 4. 900 5,300 5 020 87 2 75 7 2,970 3, 200 51 s! 210 71.4 69.6 3,330 3,240 52 4,810 4,470 77.2 64.7 ■>, 940 3,100 53 3,610 3,610 54.5 58.2 2,400 2,550 54 4,440 4,720 97.6 94.9 2,710 2,900 Shortleaf pine 55 6,400 6,690 7,610 6,880 27.0 27.1 4,340 4,500 56 28.4 26.6 4,050 4,210 57 6,670 6,990 27.0 26.4 4,100 4,340 58 7,310 7,490 28.5 26.8 4,100 4,030 White pine 101 5 070 7,200 6,890 15.4 16 2 5,410 5,720 102 6^340 ii!o 11.7 4,920 5,520 103 7,070 8,750 12.2 10.5 5,140 5,760 104 4,900 6,680 12.1 8.2 4,360 4,700 105 6,640 6,890 10.6 11.2 5,450 5,310 106 6,180 7,650 11.6 11.3 5,190 5,420 107 6,080 6, 090 11.5 11.5 4,810 5,170 108 5, 510 5, 810 11.1 10.7 5,100 4,710 109 6,930 7,300 11.4 10.5 5,330 5,080 110 5, 930 6,010 12.1 11.6 4,600 4,670 111 4,010 .1, 040 13.0 13.0 4,270 4,390 362 FORESTRY INVESTIGATIONS U. S. DEPARTMENT OF AGEICULTURE. OnSEHVATlONS AND DEDUCTIONS. («) The dift'creiice between the values for the large beam and the average for the small beams is not at all constant, either in character or quantity; the large beam may be stronger (20 i>er cent of the cases) or practically as strong — i. e., within 10 per cent (57 per cent of the cases) — or it may be weaker, and vary often considerably from the average (23 per cent of the cases). Of 6% tests on small beams 2:i,5 furnished results smaller than that of the large beam. Again, out of 390 small beams fully 40 per cent were weaker that the largo beam, while of another series of 300 only 24 per C(mt gave lower values. (h) There are in every case some siuall beams which far excel in strength the large beam; even in such cases, where the average strength of the small beams is practically the same as that of the large beam, some small beams show values 25 to 30 per cent greater than the large beam. ((') In only G per cent of the cases each of the small pieces gave a higher result than w.as obtained from the large beam, hut in these cases the latter was evidently defective. (d) In all beams the differences observed between the several small beams themselves are far greater than that between the average value of the small beams and the value of the large beam from which they are cut. From these observations, which are fully in accord witli the observations on the numerous tests of the large general series, it would ajipear that — (1) .Size alone can not account for the diflerences observed; .and, therefore, also that a small beam is not propor- tionately stronger because it is smaller, for it may be either stronger or weaker; but that if it is stronger, the cause of this lies in the fact that the larger beam contains weak as well as strong wood, besides other defects, which may or may not appear in the small stick. (2) Generally, but not always, a large timber gives values nearer the average, since it contains, naturally, a larger f|uantity as well as a greater variety of the wood of the tree; and, therefore, also — (3) Small beams, for the very reason of their smallness, containing, as they do, both a smaller (luautity and variety of the material, give results which vary more from the average than results from large beams, and. there- fore, can be utilized only if a sullicient number be tested; but it also appears that — (4) To obtain an average value, even a very moderate number of .smaller jiieces, if they fairly represent the wood of the entire stem, give fully as reliable data as values derived from a large beam. (5) Arcrai/e ralnen derived from a large xeririi of tests on small hut represenlaliee material may he ased in praetiee with perfeei safctij, and these arer((fii;s are not likehj to he modified hy tests on large material. It might be added that both the practicability and need of establishing a coefficient or ratio between results from tests on largo and small lie.ams or columns falls away. To deserve any conlidenco at all, only a large series of tests on either Large or small beams would s.atisfy the re(iuirement of establishing standard values, while a series of small pieces has the preference, not only on account of greater cheapness and convenience in establishing the values, but still more for the reason that only by the use of small, properly chosen material is it possible to obtain a sufficiently complete representation of the entire log. Before these results, part of wbicli were published by installments, had all been eomputed and arranged, the results of the work made it possible to publish, for the first time in the English language, a brief exposition of tiie teehniral properties of wood in general, which appeared as Bulletin 10 of the Division. This little booklet was copied verbatim several times by different tech- nical journals of this country, was embodied in toto in one of the best works on the materials of engineering, and was even translated into French by one of the foremost i>ublishers of France, besides being nsed itself as a text-book by several of our largest colleges. In addition to the discussions of the several technical properties of wood, tliis booklet contains the first attempt in the Engli.sh language at a key by which our common woods may be safely recognized from their structure alone. The key and some of the tables in this bulletin have been reproduced in an earlier jjart of this report. By this time, when the work was interrupted by superior orders, there were brought together the strength values for the wood of 32 species, of which 2(J were represented by more than 200 tests each (the longleaf pine by over 0,000), 17 of them by over 400 tests per species, and seven by over 1,000 tests. These results were published in full in (circular No. 1.5 of the Division, from which the following extract is here repeated: Summary or Mechanical Tests on TniUTY-TWo Si-ecies of American Woods. GENERAL REMARKS. The chief points of superiority of the data obtained in these investigations lie in, (1) Correct identification of the material, it being collected by a competent botanist iu the woods; (2) selection of representative trees with record of age, development, place and soil where grown, etc.; (3) determination of moisture conditions and specific gravity and record of position in th(^ tree of the test pieces; (4) large number of trees and of test pieces from each tree; (5) emphiynient of large .and small-sized test material from the same trees; (0) uniformity of method for an unusally large number of tests. The entire woiU of the mecdianic.al test scries, carried »n tlumigh nearly six years intermittently as funds TIMBER PHYSICS — STRENGTH OF SPECIES. 363 wero available, comprises so far 32 species with 308 test trees, furnishing over 6,000 test pieces, supplying material for 45,336 tests in all, of which 16,767 were moisture and specific gravity determinations on the test material. In addition to the material formechanical tests, about 20,000 pieces have been collected from 780 trees (including the 308 trees used in mechanical tests) lor physical examination to determine structure, character of growth, specific gravity of f^reeu and dry wood, shrinkage, moisture conditions, and otlier properties and behavior. In addition to the regular series of tests, the results ol' which .arc recorded in the subjoined tables, special series, to determine certain (|uestions were planned and carried out in part or to finish, .adding 4,325 tests to the above number. rtncoiiiil of test material. No. Name ol* species. Num- ber of trees. Number of me- clianicll tests. Average specific gravity ol'iirv woimI. Localities and number of trees from each. 1 68 12 22 32 17 8 4 20 4 6,478 2,113 1,831 3,335 540 412 696 3,396 354 225 1,009 911 256 935 299 479 222 132 649 1,035 794 300 197 100 294 172 84 91 201 476 45 508 0.61 .63 .51 .53 .38 .50 .44 .46 .37 .51 .80 .74 .80 .74 .73 .73 .72 .73 .72 .73 .81 .85 .73 .77 .78 .78 .89 .54 .74 .62 .62 .59 Alabama, coast plain (22) a; nplands(6); bill district (6) ; Georgia, undulat- ing upl.iuds (G) : Soutli Carolina, coast plain (7): Mississippi, low coast plain (2); Louisiana, low coast plain, gravelly soil i7); sandy loam (6); Texas, low coast plain (6). Alabama, coast plain (6) ; Georgia, uplands (1) ; Soutb Carolina, coast (5). Alabam.a, uplands (4) ; Missouri, low hilly uplands (6); Arkans.as, low billy uplands (6); 'I'exas, ujdaiids (0). Alabama, mountainous ].l;tii';in ih) ; low coast plain (6) ; Ark.ansas, level flood plain (5) ; Geoigi.i, lovil mast plain (6) : Soutb Carolina, low coast plain (7). Wisconsin, cl.iy uplands (5); sandy soils (4) ; sandy loam (5) ; Michigan, level drift lands (3). "Wisconsin, drift (5) ; Michigan (3). Ababama, low coast plain. South Carolina, pine barren (6) ; river bottom (4) ; Louisiana, coast plain, border of lake (4) ; Mississippi, Yazoo l>ottom (3) ; upland (3). Mississippi, low plain. (From lumlier yard.) Alab.ania, ridges of Tennessee V.alley (5) ; Mississippi, low plain (7). Mississippi, low plain (7) ; Arkansas, Mi8sissij>pi bottoms (3). Alabama, Tennessee Valley (5) : Arkansas, Mississippi bottom (3). Alilbania, Tennessee Valley (4) ; Arkansas, Mississippi l)ottonis (3) ; Missis- sippi, low ]dain (4). Alabama, Tennessee Valley (5) ; Arkan.sas, Mississippi bottom (2). ft / Arkansas, Mississippi bottom. Alabama, Tennessee Valley (5) . Mississippi, low plain (4). A]ab.ama, Tennessee Valley (5); Arkangas, Mississiiijii bottom (3); Missis- sippi, low plain (4). Alabama, Tennessee Valley (5) : Arkansas, Mississipiii bottom (3) ; Missis- sippi, low plain (3). Mississippi, alluvial plain (3); limestone (3). Mississippi, low plain. Do '> {PiDiis paluetris.) 1 (Pinna hetorophylla.) 4 (Pinus ecliinata.) R (Finns tieda.) 6 7 (Pinus atrobua.) Red pine (Pinu.s resinosa.) H (Pinus glabra.) 9 10 (Taxodium distiehum.) White cedar (CIiiini.TTyparis tbyoidea.) 11 (r.4.inb>tsuga taxilolia.) Wbity uiik . 12 10 8 11 3 5 4 12 11 6 4 o 4 3 3 3 3 1 7 T> (Qnercus alba.) v^ (Quercus lyrata.) u IS (CJuercus minor.) Cow oak (Quercus niichauxii.) 1f» (Quercus rubra.) 17 (Ouercua texana.) T* (Quercua velutina.) (Quercus nigra.) ?n (Quercus pbcUos.) ^1 (Quercus digitata.) 00 (Hicoria ovata.) 0^ (Hicoria alba.) ?4 (Hicoria aquatica.) Or^ (Hicoria minima.) 26 (Hicoria niyristicjtfomiis.) Pecan hickory • . (Hicoria pecan.) Do. Do ''fi (Hicoria glabra.) Mississippi, bottom. oq (Ulmus americana,.) :^o (Ulmus crassifolia.) White ash Misaiasippi. bottom. Do 31 (fraxinus americana.) Green ash s*> (Fraxinus lanceolata.) Sweet *'uni . Arkanaaa, bottom (3) ; Mississippi, low plain (4). (Liiiuidambar styraciflua.) a Sixteen of these were bled trees to study the etTects of boxing. & These two should probably be classed as Soutliern red oak. Thoy were collected before the distinction was finally decided npon. Note.— The values for specific gravity here given refer to "dry'' wood of test iiialerial— i. e., wood containing variable amounts of moisture below 15 per cent; the moisture Btlect has therefore not been t.aken int" :!■ fiiuiit, but more careful experiments indicate thtit its nfluence on specific gravity at such low per cent is so small that it may be neglected lor practical purposes. As will be observed, some species, notably the Southern pines, have been more fully investigated, and the results on these (wliich have been published more in detail in Circular No. 12) may be taken as authoritative. With those species of which only a small number of trees have been tested this can be claimed only within limits and in proportion to the number of tests. 364 FORESTRY INVESTIGATIONS V. S. DEPARTMENT OF AGRICULTURE. The great variation iu strength which is noticeable iu timber of the same species makes it necessary to accept with cantiou the result of a limited number of tests as reiiresintiuf; the average for the species, for it may have liappcnid that only all superior or all inferior material has been used in the tests. Hence we would not be entitled to conclude, for instance, that pignut hickory Is 14 per I'ent stronger than shagbark, as it would a)ipear iu the t.able, for the 30 test jiieces of the former uniy easily have been superior material. Only a detailed examination of the test pieces or a fuller scries of tests would enlighten us as to the comparative valne of the results. Till- following data, therefore, are not to ho considered as iu any sense hual values for the species, except where the number of trees and tests is very largo; J!e8iills of lr$l>i hi compresxinii endwise. [Pounds per square iuch.] Species. Reduced to 1:'> per cent moisture. Longleaf pine Cubau piue Shortleaf pine liOltlnlly piue Reduced to U per cent moisture. Wliite pine Ked piue Sprnee piue Billd evpress Wliile red.ar l)ouy;las spruce a White oiiU ( K'ert^up oali Post ojik Cow oak Ited o:ik 'rcxan oak Veliow oak Water oak Willow o.ak Spanish oak Stiagbark laickory Mockeruut liickoVy Water hickory iJittcriuit hickory Nut (I leg liickory Pecan liickory Pignut liickory AThiteelm '. Cedar elui White ash Greeu ash Sweet gxim Number of tests. 1,S30 410 330 660 ISO 100 170 655 87 41 218 •JIU 49 250 57 117 40 31 153 251 137 75 14 25 72 37 30 18 44 87 10 118 Highest single test. Lowest single test. 11.900 10, 600 8,500 11,200 8, 500 8,211" 10,(100 », 900 6,200 8, tlllO 12,500 ! 9,100 8, 200 11,500 9,700 I 11,300 ! 8, 600 9, 200 11.000 10,600 13,700 12, 200 10,000 11,500 12, 300 10, 500 13,000 8, 800 10, '600 9, 600 fl, 800 8,900 3, 400 2, 800 4, 500 3, 900 3, 200 4,3011 4, 400 2, 900 3,200 4, 100 5, 100 3, 700 5, 000 4,600 5. 400 5, 800 5, 500 6, 200 4, 200 3, 700 5, 800 fi. 200 0,700 7, 300 6, 400 5, 8!I0 8,700 4,900 6,200 5, 000 6, 600 4,600 Average highest 10 per cent of tests. 8,600 9, 500 7,600 8,700 6, 800 8, 100 8,8110 8, 500 6, 000 8,100 11,300 8, 600 8, 100 9, 8U0 9, 20O 9, 800 8,300 9, 000 8, 700 ' 9, .500 10, 900 11,000 9, 000 11,200 11,000 10, 400 12,700 8, 800 10, 100 8, 700 9,800 8,500 Average lowest 10 Average pdr cent I of all tests. of tests. 5, 700 6, 500 4, 800 6,400 4, 000 4, 900 5, Olio 4, 200 4,4110 4, 20(1 6, 3110 0, 000 6, 000 5, 000 5, 500 6, 900 5, 800 6, 300 5, ,500 5, 100 7, 500 8, 000 7,000 7,800 7,100 7, 300 8, 900 5, 000 0, 500 5,700 6, 000 5, 000 6,900 7,900 5, 90P 6,500 6,400 6, 700 7, 300 6, 000 5, 200 5, 700 8, ,500 7, 300 7, 100 7, 400 7, 200 8, 100 7, 300 7, 800 7, 200 7, 700 9, .500 10, 100 8.400 !l. tioo 8, 800 9, 100 10, tlOO 0, .500 8, 000 7. 200 8,000 7, 100 Proportion of tests within 10 per cent of average. Per cent. 53 61 47 49 Proportion of tests within 25 per cent of average. Per cent. 90 93 90 84 93 96 95 74 99 65 81 95 100 89 94 98 100 100 88 94 97 99 100 100 97 95 100 88 95 96 100 97 a Aotual tests on "dry" material not reduced for moisture. The v.ari.ation iu strength in wood of the virgin forest, as will be seen from the tables, is in some species so great that by projicr inspection and selection values ditt'eriug by 2v> to 50 per cent may be obtained from different parts of the same tree, and values differing 100 to 2O0 jier cent within the same species. These diflerences have all their definite reeognizaldl^ causes, to find and formulate which is the final aim of these investigations. The tests are intentionally not made on selected material (excejit to discard absolutely defective pieces), but on material as it comes from the trees, so as to arrive at an average statement for the species, when a sufficient number of trees has been tested. How urgent is the need for data of inspection as above indicated will appear from the wide range of results recorded. To enable any engineer to use the d.ata here given with duo caution and .iudgment, not only the r.anges of values and the average of all values obtained, but .also the proportion of tests which came near the average values, have been stated, as well as the aver.age results of the highest and lowest values of 10 per cent of the tests. With this informatiou and a statement of the actual number of tests involved, the comparative merit of the stated values can be Judged. With a large number of tests, to be sure, it is more likely that an average value of the sjiceies has been found. The actual test results have been rounded off to even hundreds in the fables. FACTOR.S OF SAFETY. With such lowest standard values, also lowest factors of safety could be employed. As to factors of safety, it may be proper to st.ate that the final aims of the present investigations m;iy be summed up in one proposition, namely, to establish rational factors of safety. It will be admitted by all engineers that the factors of safety as used at ]irescut can hardly be claimed to be more than guesswork. There is not an eugineer who eould give .account as to the basis upon which numeiically the factors of safety for wood have been established as "8 for steady stress; 10 for varying stress; 1,5 for shocks" (see Merriman's Testbook on the Mech'inics of Materials); or as 1 to .5 for "dead" load and ."> to 111 for "live" load (see Rankiue's Handbook of Civil Engineering). TIMBER PHYSIC8 FACTOR OP SAFETY. 365 'J'lie have developed. In the following two diagrams the average strength of the difi'erent species in compression endwise and bending, as found in the preceding tables, has been plotted with reference to the dry weight as given in preceding table. Considering that these tests and weight determinations (especially the latter) wore not carried on with thiit finesse which would be reiiuired for a scientific demonstration of a natural law, that other inlliienees, as crossgrain, unknown defects, and moisture conditions may cloud the results, and that in the averaging of results undue consid- er.ation may have been given to weaker or stronger, heavier or lighter, material, the rcl.ixation is exhibited even by this wholesale method with a remarkable degree of uniformity bordering on demonstration. An exception is apparent in tlie oaks in thai they do not exhibit this rel.ation of weight and strength with reference to other species, and also with less definiteness among the various species of oak in themselves. The structure of oak wood being exceedingly complicated and essentially diU'eicnt from that of tlie wood ..f all other species under consideration, it may reasonably be expected that it will not range itself with these. TIMBER PHYSICS STRENGTH AND WEIGHT. 367 Reaidls of testa in hcndiny, at relatiw elastic limit. [Pounds per square inch.] No Species. Reduced to 15 per cent moisture. Longleaf pino Ouban pine Sbortleaf pine Loblolly pine Jieduced to 12 per cent moisture. White pine Ited pine Spruce pine li;ild cypress White cedar l>ou;L;las sjirucetr White oak Overcu}! oak Post oak Cow oak Ked oak Texan oak Yellow oak Water oak Willow oak Spanish oak Sbagbark hickory Mi>ckernut hickory Water hickory Bitternut hickory ^Nutniej; hickory". Pecan hickory Pignut hickory White elm Cedar elm White ash . ^ Green ash Sweet gum , Number of tests. 1,100 a9o ■SM 650 i:iO 9.5 170 055 87 41 •JIS •_'!« 4!) •J5li 57 117 40 :n 153 257 187 75 14 25 72 37 30 18 44 87 10 118 Highest single test. 13, 500 12, '.too 11,900 12, 700 10, 000 11,300 13, 700 12, 000 8,200 13,700 15, 700 11,000 lO.liOO 14, 2U0 14,500 12, UOO 11, 800 11, 800 13, 100 13, 500 16, 100 15, 400 11,900 14, 300 12, 200 15,1100 17,500 9,700 10, 700 11, 500 13, 200 11,000 Lowest aiugle test. 2, 400 2, 200 2,900 3,100 4,100 3, 100 3,000 2,200 3.400 2,800 4,400 4, 000 5,100 3.400 5, 100 5,900 4, 900 4, 500 2, 700 5,100 5,400 4,300 4,100 7,500 4, 200 5,800 7,400 5, 300 4, 700 3,600 3,200 3,500 Average of high- est 10 per cent of tests. 11,100 11,500 9,700 10, 800 8.200 10, 300 11.200 9, 900 7, 390 9, 600 14, 100 9,500 9,600 11, 600 13, COO 11, 400 11,100 11, 400 10, 000 11,600 14, 200 14, 600 11,800 14, 000 11,200 14,400 16 400 9, 600 10, 100 10, 400 13, 200 10, 100 Average of lowest 10 per c.ent of tests. 5,400 5,600 4, 800 5,400 4, 500 4,500 5, 000 4, 200 4,000 3, 400 0, 100 5,400 6,000 5, 000 5,600 7,800 5,100 5, 500 4, 300 6, 600 7,700 7,800 4, 800 7,600 6,400 7,900 8, 300 5, 400 5,800 5,200 3,200 5,100 Average of all tests. 8, 500 9,500 7,200 8,200 6,400 7,700 8.400 6, BOO 5,800 6, 400 9,000 7, 500 8,400 7, 600 9, 200 9,400 8,100 8,800 7,400 8,600 11,200 11, 700 9, 800 11,100 9, 300 11,. 500 12,600 7, 300 8,000 7,900 8,900 7,800 Proportion of tests within 10 per cent of average. Proportion of testa within 25 per cent of average. Per cent. 43 42 48 46 Tcr cent. 81 83 81 85 Alodulua of elasticity (average of all tests) . 1, 890, 000 2, 300, 000 I, 600, 000 1, 950, 000 300, 000 620, 000 640, 000 290, 000 910,000 6K0, 000 090, 000 620, 000 030, 000 610,000 970, 000 860, 000 740, 000 000, 000 750, 000 930, 000 390, 000 320,000 080, 000 280, 000 940, 000 530, 900 730, 000 540, 000 700, 000 640, 000 050, 000 TOO, 000 a Actual tests on " dry " material not reduced for moisture. /siipao //,000 i moo 9000 'i. 8000 17000 6000 SOOO Weight pur ctibii; foot in pounds. Fig. 95.— Eelation of strength in compression endwise to weight of material. The ligurc at each point indicates the species thereby represented. 368 FORESTRY INVESTIGATIONS U. S. DEPARTMENT OF AGRICULTURE. moo /ROOO /7000 moo /sooo uooo I /3,000 /2000 f //ooo /oooo 9000 8000 7000 ^000 AVeijiht per cubic foot in i>oimds Fui. 96.— Kelaliou of weight to bending strength at rupture. The figure at each point iutUcatea the epeciea thereby represented. TIMBER PHYSICS UNIFORMITY OP STRENGTH. 369 In addition, the difficulty of seasoning oak without defects or even securing perfect material may have influenced the results of tests so as to cloud the relationship with the genus. If further close study, su|)pli-mented by additional series of tests carefully devised to investigate this relation- ship, should uphold the troth of it, this result may ho set down as the most important practical oue that could he reached by these tests, for it would at once give into the hands of the wood consumer a means of determining the relative value of his material as to strength and all allied properties by a simple process of weighing the dry material ; of course with due regard to the other disturbing factors like crossgrain, defects, coarseness of grain, etc. Ri'snits of (<•»(•< ill aiiiiiire.K.th>ii iivioss iirain (n) and shearing u-ith grain. [PouBds per square inch.] Species. Iteduced to 15 per cent moiittire. Loiigleaf pine, f'libanpine .-. Shortleal" ]»iue. Loblolly pine. - Reduced to IL' per cent moisture. White pine Red pine Spruce i>ine Bald cypreea While'cedar Doiiiilas 8pruc©6. AVhiteoak Overeup oak Po.st oak Cow oak Red oak Num- ber of t«sts. Compres- Bion across grain. 1,210 400 330 690 130 100 175 650 87 41 218 216 49 256 57 Shearing ! with 1 1 grain not 'jj reduced for moisture. 1,000 1,000 900 1,000 700 1,000 1.200 8110 700 800 2, 200 1.900 3, out) 1,900 2,300 700 700 700 700 400 500 800 .WO 400 500 1,000 1,000 1,100 900 1,100 If. 17 1 18 19 20 21 1 22 1 23 24 25 26 27 28 29 30 31 32 Species. Redxiced to 12 per cent moisture — Continued. Southern red oak . . . Black oak "Water oak "Willow oak Spanish oak Sliagback hickory. "White liiflvory — "Water hickory Bitternut hickory. Nutmeg hickory -. Pecan hickory Pignut hickory White ehu Cedar elra "White ash Green ash Sweet gum Num- jer of tests. Compres- sion across grain. 117 40 30 153 255 135 75 14 23 72 37 30 18 44 87 10 118 2,000 1,800 2.000 1,600 1,800 2,700 3,100 2,400 2. 200 2.700 2,800 3. 200 1,200 2,100 1,900 1,700 1,400 Shearing with grain not reduced for moisture. 900 1,100 1,100 900 900 1,100 1,100 1,000 1,000 1,100 1,200 1,200 800 1,300 1, lOO 1,000 800 oTo an indentation of 3 per cent of the height of the specimen. b Actual tests on "dry " material not reduced for moisture. Having fully established tlie great influence of moisture on the strength of wood, the practi- tioner still needed information as to the rate and manner of drying and as to the way in which moisture is distributed during seasoning. Several thousand moisture determinations were made and it was established beyond doubt that moisture is generally least abundant at the ends, is quite evenly distributed throughout the length, but i.s not always uniform in different parts of the .«ame cross section, often varying in this respect within astonishing ranges, so that the use of timber in a half-seasoned condition, and where uniform seasoning can not be obtained by the material, requires that these facts be duly considered in designing. Tests of Maximum Uniformity. Both in this country and abroad small differences in strength values were often interpreted as deciding for or against any given material. This same problem arose also in every case where many results were to be compiled, and it seemed especially desirable once for all to find just how much uniformity could be expected of wood materials. From a large series of well-selected quarter-sawed pieces repre.senting several kinds of pine, cypress, and hardwoods it was found that even contiguous blocks, 2.J inches long, may differ by as much as 2 to 4 per cent in conifers and as much as 13 per cent in oak, and that in a scantling only (i feet long the butt might differ from tbe top by 10 to 20 per cent in conifers and over 35 per cent in oak. This extremely valuable set of results throws much light upon discussions of the past, and is well suited to show that many boastful claims rested on very flimsy and entirely unreliable differences, such as might well be accounted for by a little more extended examination of materials. It will also assist in judging test results in the future and help to avoid useless controversy and prejudice. The following more lully illustrates the results of this series: Scantlings of air-dry material, 6 to 10 feet long, of white pine, longleaf pine, tuliptree (poplar), and white oak, and of perfecUy green material of loliloUy pine and cypress, fresh from the saw, were cut partly into blocks 2 by 2 by 2i inches, bur mostly into cubes of 2| inches. All material was quarter sawed, carefully prepared, and in all cases treated alike, either perfectly green or dried together at the same temperalnre. Altogether 529 tests in endwise compression were made, namely, 100 ou white pine, 72 on longleaf pine, 99 on loblolly pine, 10 on white oak, 115 on tuliptree (poplar), 103 ou cypress. H. ])oc. 181 24 370 FORESTRY INVESTIGATIONS TT. S. DEPARTMENT OF AGRICtJLTURE. From these tests the following table of averages is ilcrivnd, together with tig. !I7: Average of testa for maximum uniformity. Namir. "White pine (Finns strohna) Longleaf pine (Finns palnslris) Tuliptree (poplar) (Liriodendron tnlipifera) White oak (t^nerene allia) Lohlolly ]iiue (Finns t.;eda) Cypress (Taxodinm distiehnm) Moisture, Per cnlt. 8 7.8 8 Yard dry. 125 + (green). 125 + (green). Average strength of all pieces. (Ireatest difference in strength hetween ale conditions of a lieam of rectangular cross section loaded at the middle. Regarding the distribution of internal stresses, it must be agreed that the neutral plane lies in the center of the beam so long as the beam is loaded within the elastic limit ; this follows from the fact that the modulus of elasticity is the same whether derived from com.preBsion tests or from tension tests (i. e., Ec = Et), as proved by experiments of Niirdlinger, Bauschiuger, Tetmayer, and others. Since the distortion of any given tiber in the beam is proportional to its distance from the neutral plane, the distriluition of stresses in a longitudinal section of .a beam loaded up to its elastic limit may be represented by the followiug diagram, iu which the vertical scale represents increments of distortion and the horizontal scale the liber stresses. In this diagram the angle a = angle h, since E<. = Et ; ami furthermore, since these latter quantities are each equal to the modulus of elasticity obtained from cross-bending tests (.according to the same authorities), this angle a (or h) van be obtained by platting the results of the cross-bendiiig test itself. It is a wellestablished fact that the tension strength of wood is much greater than the compression strength, ,an p. ^ ^^ \ . ] K \ 1 P - \ 1 k \ 1 ^8 - \ ; ^/<. 1 1 1 \i/ 1 1 Z. ^2 3 4 S 6 Fiti. 98.— Relation or fiber stresses and diatortion:*. \/y. STfffSSES //V WOO LBS. /234SS789/0// 'e , , ^ 1 1 1 1 1 1 1 1 7 *2 --' -"' r NEUTRAL AX/S 1 1 1 1 r 1 1 1 U.-P^~T-~^i/l 8 L / 2 3 ^ S 6 7 3 ^ /Oy/ Flo. 99.— Distribution of internal stresses iu a beam at rupture. Experience iu testing, however, has taught that when a piece of green wood is tested in ^ompre88ion it will undergo a great distortion after the maximum load has been applied without actually biealcing down — in fact, while sustaining the same load. A piece tested in tension, on the other hand, breaks suddenly as -ioon is the maximum load is applied. A beam iu failing may, therefore, sustain an increasing load long after the extreme compression fiber has been loaded to its ultimate strength; the fibers on the compression side continue to be mashed down, while the ueutral plane is lowered and the stress in the tension filter increases until, very often in practice, the beam "fails iu tension." With these facts and ^ rO/RC£S L 2 3 4 S //V S 7 WOO 8 9/0 LBS. // /a observations before us it is possible to con- struct a diagram so that it will represent, approximately, at least, the distribution of internal stresses iu a beam at rupture. (See fig. 100.) In this figure OA represents the position of ueutral plane at time of rupture, OU the distortion in the extreme compression fiber, UC the stress on same fiber, OL the distor- tion in extreme tension fiber, and LT the stress on that fiber. It can readily bo seen that the manner of breaking will influence slightly the form of this diagram. If the beam falls in com- pression before the tension fiber reaches its elastic limit the line OT will be straight as shown, otherwise the line will assume some such positiou as Oi,T, (diagram 99), iu which /, is the elastic limit iu tension. From the approximate distribution of internal stresses their relation to the external load may be determined. The two funda- mental equations — (1) that the sum of inter- nal stresses on the tension side equals the sum of internal stresses on the compression side, and (2) that the sum of the external moments equals the sum of the inter- nal moments — apply at the time of rupture as well as at the elastic limit. From (1) it follows that area OUC?=^area OLT, and the jtosition of the neutral plane at rupture is thereby fixed. If now the line LU be assumed to represent the depth of the beam iu inches instead of indicating the distortion of the fibers, the sum of the internal moments about the point O is found by multiplying the area of either the compression or tension diagram by the sum of the distances of their respective centers of gravity from the neutral plane. By putting this sum equal to the motiient of the external load about the same point O the first relation is established. Fig. 100. — Position of neutral axis and internal stresses at rupture of beaui. 374 FORESTRY INVESTIGATIONS U. S. DEPARTMENT OF AGRICULTURE. RELATION or CRUSHINU-KNDWISE STRKNOTU. The secoud relation (that of crushiiiK-endwise strenstli to iutonial stressos) was touched upon iu discuasinj; the first, whiu it was stated: (1) That the true elastic limit of the beam is j)i()lial)ly rcachud at the same instant that the extreme libers on the compression side reach their elastic limit in cDmprcssiou. (2) That this latter limit lies close to the ultimate compression-eud wise streu^tli (so close that former experimenters have been unable satisfactorily to separate them). (:i) That a piece of gieen wooil will stand a ^rcat deal of distortion after the ultimate load is applied before actually failing. And to these statements may be added the evident fact (4) that the stress on any fiber on the (•onipressiou side can not exceed the compression-endwise strength of the material. (5) Finally and most im])ortant it appears from (1) and (2), but especially from an examination of the several thousand test results on tlie several species of conil'crs made by the Division of Forestry, that the extreme fiber stress at the true elastic limit of a beam is practically identical with the compression-endwise strengtii of the uuiterial. (This last observa- tion, which was forced upon the writer by its continual repetition in the largo series of tests under review, lies at the basis of this discnssion.) The observation of this identity makes the distribution of internal stresses appear more simple than was hitherto assumed, and the desired relation between compression and cross-bending strength capal)le of matlu'matieal expression. DEVELOPMENT Of FORIIUL.E. From these considerations the distance UC in fig. 100. which represents the ultimate compression-endwise strength of the material, becomes practically equal to the distance el, which represents the compression strength at the true elastic limit, and hence the line IC straight and vertical; and if OT is taken as straight, the diagram will be made up of simple geometric figures, as in iig. 100. The line LU will represent the total fiber distortion at time of rupture, and is equal to the sum of the amounts by which the extreme compression fibers shorten and the extreme tension fillers elongate. Let a test in which the following (luantities have been observed and recorded be considered : Let Pr= the external load at rupture (pounds). ^ ^r^the corresponding defiection of the beam (inches). C = compression-endwise strength of the material (jiounds). E^ modulus of elasticity (pounds). rf=:de])th of beam (inches). 6 = brcadth of beam (inches), i^ length of beam (inches). zJe^detlection at true elastic limit. Then, based upon the above statements, by means of formulas derived from the geometric relations of the diagram and the fundamental eijuations of equilibrium, the following ([uantities can be eaUiilated: Let t',,=; total liber distortion due to bending at true elastic limit (inches). Er-i= total fiber distortion due to bending at rupture ^:Llj (inches). (Jp^ distortion in extreme tension fiber at rupture = LO (inches); also the proportional dis- tance of neutral plane from tension side of beam. (Jr^real =;^OC4'-'ca Distancf from net tral plan of center gravity CQ'«*-*'i'-»??"*'^-1'-i"-*"*'^ •■B^xe aoib-no^ jo S3S£SS;SSSS2?2S5B eo«Nco'c^corteoMcoC'ieiim ili 'dptS lI0I6fJ9Jdui09 UQ d i 23, 300 22, 760 27, 400 24, 000 25,010 24,900 22, 600 27, 600 30, 500 18, 000 22, 200 28, 100 23, 600 •gpiB iinit^uo) U() H 23, 400 23, 053 27, 100 24, 000 24, 800 24. 600 22, "00 27, 600 30, 500 18, 000 22, 500 28, 100 23, 900 ■jaqg aoieua^ (»ui9JiX9 jo dJiudnj jb bbsjis T Lbs. per sq.io. 9,700 9,810 12, 200 9, 800 9. 920 9,820 8,350 11, 300 11,500 6,780 11,000 11,330 10,400 Real dis- tance of neutral plane at rupture. isnl' yirq 1[0!11a\. op!s uoi9 -s^jdoio.) «o J9qi| ^tiq^ oioj^j -c [ P3 M fh ,-. cj f i c^i — c-i ci ^ f rH nreaq jo 9pi« aojsttgi uioj^ ■« .^.^^^iftuiin^'uSif. rfT^^ •.».innliu jB .umy uoiKU^i kiiiinj)X9 ui uin)JO(«!(i ■0' -,■ t~ oo -.D ,0 cc .r eo in =: r- t^ o=>oo-,c 000=0^ = 0000000000000 oooo'oo'ooodo"dd VJUUJI «• %J Total fiber dis- tortion due to bending. •gju^dnj %Y 0. O^QOO-t?liftCDOr^O Mffl.*tOU5lOOt~iO.-i-^QOO 0000000000000 "\\m\\ o\%9V\9 lY H 0039 0055 0059 0054 0053 0058 0046 0055 0062 050 0059 0060 0050 1 0000000000000 'inmioi'iv '«[9 enj) IB qiSa9J'}e Satpaag^ '^ »3 e^'v-ii''^'-*'^'"'^ ..-r&i©rwi05 •a.nil(luj iv, (J (iiind jnoqu 8|U3mom iBu-iajai jo rang a ! ="50 •!»«o*r.f •T*;ii«*»»»'f 1 a --> a 1 'Ojn^duj IB itaiod )uoqii s!)a9uioui [Biud^xd j6 nine xitn^ay 51" ''\\01l\ Ot^SBp eUJ') fB aOI)3909(£ t~ •|!tu(i oijeB(:t ana; -jir pBOi p. 3 gg=go=gS§g2§S o.-c 0000 O' 000000 ?5C03ir-t~*■» in ".jfrot~io« •-V'jini'jeRii* JO 8innpop[ ,, Og, — TM-f^^ — — — il-f-'SiX' •iunidnj ^u pBOT ^ a 1 88ggs§S3SSggg 1 oinco^ooift ■^■Mtct-ooifto 1 N (M rt (M M C) M 5-1 .X r- .-^ CI •qiSti.uiB Snipuag « So 3« 000000=000000 tot~o-*w — m^Hloo^~■T»■^-. rttoco■^fl«p■v-ft~x^clrt(Dco l-(Doot-t~r-c~t— aom0 4,. 540 4,470 3,000 4,'JSO 3,600 3,760 ! 4,708 5.3 lO 5,057 4,-J03 4,571 4.4-JO 4,578 4,169 3.854 3,3 1-i 3.«1»7 3,760 4,430 3,969 4,-2-20 4,296 4,129 4,178 3,078 3.860 3,8»S 3,395 1.46 1.56 1.48 1.45 1.49 1.47 1.4S 1.47 1.47 1.44 1.45 1.23 1.31 1.24 1.-22 1.25 1.23 1. 25 1.23 1.23 1.21 1.22 10.517 10. 979 10.885 9, 675 9,894 9,943 9,104 9,274 8,796 8,926 8,415 7,677 8,564 8, 1155 7,014 7,371 7, 30« 7,381 6, 810 6,465 6,427 0,101 7,719 8,552 8,026 7,061 7,376 7,290 6, 840 6.751 6, 403 6,485 6,124 0.97 1.04 99 0.97 0.99 0.98 0.99 0.98 0.98 0.96 0.97 1.18 1.26 1.19 1. 17 1.20 1. 18 1.20 1.18 1.18 0.87 1.17 58,760 66, 380 63, 216 52, 535 .57, 144 .55, 248 57. 222 52,118 48, 177 ' 41, 400 46,219 2,200 2,800 2,400 2,400 2, 600 2, 400 2, 500 1,800 2,200 2, -200 1,940 0.296 0.391 0.413 0.345 0. 356 0.431 0.440 0. 328 0.387 0.372 0.300 a Failed, due to knot. Note.— Columns of figures in same distinctive type to be compared one with the otlur. On the whole, it is in no way boastful to assert tliat tbis work has already fnruisbed prac- tical data euough to uiore than pay the expenses incurred ten times over; that its fruits are uot half jiatbered, and that Ibr more than a quarter of a century its results will .serve as a basis for the user of wood and as the guide to the teacher and e.xperiiueuter. Devklopment of the Science oy Timber 1'iiysics and Methods Employed in the Investigation. Since the elaborate plan and methods of this study of our woods denotes an entirely new departure in timber investigations, at least in our (-ountry, it is only fitting to place the credit for its conception, for the elaboration of the plan, the organization of the work, and the persistent prosecution of the same in spite of many drawbaidvs and lack of support. This credit belongs to Dr. B. E. Fernow, chief of the Division of Forestry. The plan was first foreshadowed in his second report (1887, p. 37) as cliicf of that division, ami the word "timber physics" was there used for the first time, and the essentials of the future plan were there discussed. In a small tentative manner the first steps to put it in operation were made in 1888. In the report for 188!) we read : The investigations into the technology of our timbers and especially into the conditions ni>on which the .lual- ities of our timbers depend-for which Mr. Roth of Ann Arbor has begun preliminary studies— has also made but slow progress for lack of means. In the report for 1890 we find, besides an account of tlie tests on Northern and Southern oaks referred to before, the statement that "by the increase of ai>propiiations the forest technological investigations referred to in former reports have become possible on a scale which was hitherto unattainable," and a description of the plans is given. But the first fuller statement of the 378 KOKESTRY INVESTIGATIONS U. S. DEPARTMENT OF AGRICULTURE. (lev'eloi)iuent of the investigation and its methods was not published until 1892, in Bulletin (i, in which Mr. Fernow described the aims, objects, and methods at length. In the report for ISOO the following language is used: TIMBER TESTS. AVliild till' 1180 of woiiil i)ulii mill Dtlier substitutes mi;y displ.icd in many w.ays tlie use of wood iu its natural state, tliei'o will always be desirable i|uali(ies inlierent iu tbe latter tbat make its use indispensable. Hence the desirability of knowing the qualities of our timbers and, if possible, of kuowinj; the conditions under which the wood (lop will de%'elop the desirable qualities. iMuch work and useful work is done in the world by the rule of thumb. All such work is not reliable and certainly not ecouoniical. With the need of f;reater economy in jiroduction, the need of more accurate measuring arises, aud with that the need of more specitic knowledge of the itiaterials to be measured. Wood is one of the materials which has been measured by the rule of thumb longer than others. Iron aud other metals used in the arts have their properties much more accurately determined than wood material. Especially in the United States, when we speak of quality of our timbers, it can only be in general terms; we lack definite data. One difticulty in determining reliably the ijualities of our timbers lies iu the fact that living things are rarely precisely alike. Every tree differs from every other tree, and the material taken from the one has a different value from that taken from the other of the same species. Yet every tree has some characteristics in common with all those grown under similar conditions. But even these common properties differ iu degree in different individuals. Iii(liviroducing quality in trees, to acquire or at least enlarge this knowledge mu.st be one of the iirst and most desirable undrrtakings iu which this Division can engage. Accordingly a comprehensive plan has been put Into o|icration to study systematically our more ini|portaut timber trees. It will at once bi' understood that as loug as the qualities are to be refeired to the conditions under which the tree is grown, the collection of the study nuiterial must be made with the greatest care, and the material must be accompanied with an exhaustive description of these conditions. Since, further, so much individual variation seems to exist in trees grown under seemingly the same conditions, a largi^ number UMist be studied in order to .arrive .it reliable averiige values. For the present it has been decided to study tlie pines, especially the white pine and the three .Southern lumber pines. In selecting localities for collecting specimens, a distinction is made between station and site. By station is understood a section of country (or any plaies within that section) which is characterized in a general way by sindlar climatic conditions .and* geological forniiition. Station, then, refers mainly to the general geographical situation. Site refers to the local conditions and surroundings within the st.atiou, such as difference of elevation, of exposure, of physical properties aud depth of the soil, nature of subsoil, and forest conditions, such as mix(hj'sic8, in short, is to furnish all necessary knowledge of the rational application of wood in the arts, and at the 8aml^ time, by retrospection, such knowledge will enable us to produce in our owu forest growth qualities of given character. 380 FORESTRY INVESTIGATIONS V. .s. DEPARTMENT UF A(iKI(;UI.TUKE. Conceived in this manner it becomes the pivotal science of the art of forestry, around whicli the practice both of the consumer ;ind producer of forest growth moves. The lirst part of our science would requiro a study into gross and miuute anatomy, the structure of the wood, foiui, dimensions, distribution, and arraniieineut of its cell elements and of groups of structural parts, not only in order to distiugnish the diHerent woods, but also to furnish the basis for an explanation of their physical and mechanical i)roperti<^s. WencxI would class here all investigations into the physical nature or properties of the wood material, which necessarily also involves an investigation into the change of these properties uiuler varying conditions and inlluences. A third chapter would occupy itself with tin- chemical composition and properties of woods and their changes in the natural process of life, which jiredicato the fuel value and dural)ility as well as the use of the wood in chemical technology. Although, philosophically speaking, it would hardly seem admissible to distinguish between physical and iiieclianical properties or to speak of " mechanical " forces, for the sake of convenience and practical purposes it is disiiable to make the distinction and to classify all phenomena and changes of nonliving bodies, or bodies without reference to life functions, into chemical, physical, and mechanical phenomeua and changes. As chemical phenomena or changes, and therefore also conditions or jiroperties, we class, then, those which have reference to atomic struc- ture; as physical phenomena, changes, and properties those which refer to and depend on molecular arrangement, and as mechanical (molar) changes and properties those which concern the masses of bodies, as exhibited under the inlluenco of external forces, without altering their physical or chemical {'onstitntion. There is no doubt that this division is somewhat forced, since not only most or all mechanical (as here conceived) changes are accompanied or preceded by certain alterations of the interior molecular arraugeiuent of the mass, but also many i>hysical phenomeua or properties, like density, weight, shrinkage, having reference to the mass, might lie classed as mechanical; yet if we conceive that physical phenomena are always concerned with the "<|uantity of matter in molecular arrangement" and with the changes produced by interior forces, while the latter are concerned lather with the "position of matter in nndecular arrangement " and with changes under application of exterior forces, the distinction assumes a practical value. Our conception of these distinctions will be aided if we refer to the physical laboratory as furnishing the evidence of physical phenomena and to the mechanical hiboratory as furnishing evidence of mechanical iiheuomena. These latter, then, form the subject of our second or dynamic part of timber physics, which cimcerns itself to ascertain mainly by experiment, called tests, under application of the laws of elasticity, the strength of the material and other properties which are exhibited as reactions to the influence of applied stresses, and those which need consiileralion in the mechanical use of the material in the various arts. Having investigated the material in its normal condition, we would necessarily come to a consideration of such physical and chemical conditions of the material as are abnormal and known as disease, decay, or defects. Finally, having determined the properties and their changes as exhibited in material produced under changing conditions or dift'ering in physical and structural respects, it would remain the crowning success and goal of this science to relate mechanical and physical properties with anatomical and physiological development of the wood substance. The subject-matter comprised in this branch of applied natural science, then, may be brought into the following schematic view : I. — Wood sirih'Tike ok xyi.otomv. (a) Exterior form. Here would be described the form development of timber in the standing tree, ditferentiated into root system, root collar, boh' or (runk crown, branches, twigs; relative amounts of material furnished by each. (h) Interior utruclural appearance; diti'ereutiation and arrangement of groups of structural elements. Here would be described the gross structural features of the wood, the distribution and size of medul- lary rays, vessels, fibro-vascnlar bundles, as exhibited to the naked eye or under the magnifying glass on tangential, radial, and transverse sections; the appearance of the annual rings, their size, regularity, dif- ferentiation into summer and spring wood, and all distinguishing features due to the arrangement and proportion of the tissues composing the wood. (c) Minute nnatomij or liistologij; diffcientiation and arrangement of structural elements. Here the revelations of the microscope are recorded, especially the form, dimensions, and structure of the ditVerent kinds of cells, their arrangement, proportion, and relative importance in the resulting tissues. {(i) Comparitlire elaanificaiion of woods, according to siriicliiral features. ^c) Lawn of wood i/rowtk with reference to structural results. Discussion of the faitors that influence the formation of wood in the standing tree. (/) Abnormal forynationa. Burls, bird's eye, curly, wavy, and other structural abnormities and their causes. II. — I'liYsiCAi. I'ltorEUTlES, i. e., properties based on molecular ([ihysical) constitution. (a) Exterior appearanee. Such properties as can be observed through the unaided senses, as color, gloss, grain, texture, smell, resonance. (fc) yiaterial condition. Such properties or changes as are determined by measurements, as density or weight, water contents and their distribution, volume, and its changes by shrinkage and swelling. TIMBER PHYSICS — EARLIER WORK. 381 (c) aa^sifcaiiov of u-oods according t., physico-iechnical properties, i. c, such physical properties as determine their application in the arts. III.— Chemical i-hoi'krties, i. e., properties based on atomic (chemical) constitution. («w;,w»'ai e/.«»iica? aim/i/sis o/- »o«d (qualitative and quantitative). Here ^vould be discussed the chemical constitution of dirterent vvoods and difterent parts of trees and their changes due to physiological processes, age, conditions of growth, etc. (h) Carhohudrahx of the wood. . ■ , ^ i 4.i,„;, Here would he more specially discussed cellulose and lignin, .ork formations, organic contents and their changes, and such properties as predicate the fuel value of woods, their mannfactnre into charcoal, their food value, pulping qualities, et(\ (c) Exiractife material.i. ■ „ .,1 A knowledge of these underlies the application of wood in the manufacture oi tan extrarts, res.n, and turpentine, tar, gas, alcohol, acids, vanillin, etc. (d) '^''*'2^^'^J^"^'^2T^ ;';. jj^^^^^ ehemical properties whi,-h predicate durability and underlie processes of increasing the same, fe^ Mineral consliliients, • , i.j. * A knowledge of the.^e in particnl.ar will establish the relation of wood growth to mineral constituents of the soil andllso serve .as basis for certain technical uses (potash). IV.-MEC..ANICAL ruoPERTlES, i. e., properties base.l on elastic conditions exhibite.l by the aggregate mass under influence of exterior (mechanical) forces, (a) /■ormc;m«flC8ir«/,OH/c»/'«si«», commonly culled elasticity, tlexib.hty. toughness. (A) Form chanflc, mtU de^lr,wlio« of cohesion, conunonly ,alled strength (tensile, eom,.ress.ve, torsional, shearing), cleavability, hardness. V.—Technical PROPERTIES, i. c., properties in combination. , . , ,, . ,■ ,■ • 41 „.t„ Here would be considered the woods with reference to their technical use, their application ,u the arts, which is invariably based upon a combination of several physical or mechanical properties. VI Diseases and faults. , . , , i-.- Here would be treated the changes in structure and properties from the normal to abnormal conditions, due to influences acting upon the tree during its life or upon the timber during its use. VII —Relation of propkhties to each other. , • , , 1 Here would be discussed the connection which may be established between structure, physical, chemical .and mechanical i.roperties, and als.. between these and the conditions of growth under which the material was produced. The philosophy of the entire preceding knowledge would here be brought together. To contribute toward this important branch of human knowledge and to help in the building ol its foundation, the work undertaken by the Division of Forestry described in this bulletin was designed by the writer; and, in order to build with a kn'owledge of what has been done before on this structure, a brief review „t the progress ,n the development of timber physics seemed advisable. This historical review is then given. From this we deem it appropriate to quote the portion which refers to efforts in the United States up to the time of the writing to cstabhsh data regarding the mechanical properties of our timber: AMERICAN WORK. While it may bo possible to work out the general laws of relation between physical and mechanical properties on material of European origin, for practical purposes we can not rely upon any other data than those ascertained f 01 A 1 rican timb rs, and so .ar .as dependence of quality on conditions of growth are concerned this truth is.us as patent Although in the United States probably more timber has been and is being used than in any other country but little work has been done iu the domain of timber physics. •. , ti o ,.f An oiig the earliest American experiments falling in the domain of timber physics may be cited those of Marcus Bull to deteruune "the comparative quantities of heat evolved in the combustion of '^.'^ P""'"^! I'^"'"!! of wood and coal used in the United States for fuel," made in the years 1823 to 1825 and published in 18^6 H the experiments of Lavoisier, Crawford and Dalton, and Count Rumford on similar lines are discussed and followed by an able series of experiments and discussion on American woods and coals The only comprehensive work in timber physics ever undertaken on American timbers is that °"I'- T- P- Sharpies in conueition with the Tenth Census, and published in 1884, Vol. IX, on the Forests ot North America. ?on p hen iveness, however, has been sought rather in trying to bring under examination all the arborescent species Snirnilhing fuller data of practical applicability on those from which the bulk of our useful material i IcHv d ''The results obtained," the author says, ".are highly suggestive; they must not, however, be eonsidered conclusive, but rather valuable as indicating what lines of research should be followed in a more thorough study oi ""' NoiTess'than 412 species were examined in over 1,200 specimens. The results .are given in five tables, besides four comparative tabled of range, relative values, averages, etc. The specimens were taken " m most cases from the b °t7cut and free from sap and knots;" the locality and soil from which the tree came are given in most cases and in some its diameter an] layers of heart .and sapwood; determinations were made of specihc gravity, mineral ash per cent, and from these data fuel values were calculated. 382 FORESTRY INVESTIGATIONS U. S. DEPARTMENT OF AGRICULTURE. Tlio spoiimeiis tested were "carernlly seasoned." For transverse straiu tliey were niailo 1 ccntimoters (1.57 inches) siniare, and a i'ew of donblo lliese dimensions, with 1 meter (3.28 feet) span. One tabh> illnstrates " the relation between the specihe gravity and the transverse stren-^th of the wood of speeics, npon whieh a sufheient number of tests has been made to render suth a comparison valuable." This table seems to show that in perfect specimens weight and strength stand in close relation. A few tanning determinations on the bark of a few species are also given. The object of the work as stated, namely, to be suggestive of a more thorough study of the subject, has certainly been fully and creditably attained. Of compilatory works, for use in practice and for reference, the following, published in the United States, may be cited: Do Volson Wood: Resistance of Materials (1871), containing rather scanty references to the work of Chcvandier and'Wertheim. R. G. Hatfield: Theory of Transverse Strain (1877), which, besides other references, contains also twenty-three tables of the author's own test on white pine, Georgia pine, hemlock, spruce, white ash, and black locust, on sticks 1 l)y I inch bv l.fi feet in length. William II. Burr: The Elasticity and Resistance of Materials of Engineering, third edition, 18110, a compre- hensive work, in whieh many relerenccs are made to the work of various American experimenters. Gaetano Lanza, in .\p|ilied Mechanics, 1885, lays especial stress on the fact that tests on small select pieces liive too high values, and quotes tlu' following experiments on Iimg jiieces. He rcd'ers to the work of Capt. T. .1. Rodman, United States Army, published in Ordnance Manual, who used test pieces '21 by 5| inches and :"> feet length, without giving any reference to density or other facts concerning the wood; and to Col. Laidley's United States Navy test (Senate Kx. Doc. lli. Forty-seventh Congress, first scission, 1881), who conducted a series of experiments on Pacific slope timbers, "white and' yellow pine," 12 feet long and 4 to 5 by 11 to 12 inches square, giving also account of density and average width of rings. Lastly, the author's own experiments, made at the Watertown Arsenal for the Boston Manufacturers' Mutual Fire Insurance Company, on the colunmar strength of " yellow pine" and white oak, 12 feet long and 6 to 10 inches thick, are brought in support of the claim that such tests show less than half the unit strength of those on small pieces. Data as to density, moisture, or life history of the specimens are everywhere lacking. R. H. Thurston, Materials of Engineering, 1882, contains, jierhaps, more than any other American work on the subject, devoting, in Chapters II and HI, 117 [lages to timber and its strength, and in the cha]itci' on Fuel several jiages to wood and charcoal, and the products of distillation. 'It .ilso gives a description of some twenty-five kinds of American anil of a few foreign timlier trees, with a descrijition of the structure .ind their wood in general; directions for felling and seasoning; discusses briefly shrinkage, characteristics of good timber, the induenco of .soil and climate on trees and their wood, and of the various tonus of decay of timlier, methods of preservation and adaptation of various woods for various uses, much in the same manner as Rankine's Manual of Civil Engineering from which many conclusions are adopted. The author refers, besides foreign authorities, to the following American investigators : G. H. Corliss (unpublished?) is quoted as claiming that proper seasoning of hickory wood increases its strength by 15 per cent. R. G. Hatfield is credited with some of the best experiments on shearing strength, published in the American House Carpenter. Prof. G. Lanza's experiments are largely reproduced, also Trautwine's on shearing, and some of the author's own work on California spruce, Oregon pine, and others, especially in torsiim, with a specially constructed machine, an interesting plate of strain diagrams accompanying the discussion. In connection with the discussion by the author on the influence of prolonged stress, there is quoted as one of the older investigators, Herman Haupt, whose results on yellow pine were published in 1871 (Bridge Construction). Experiments at the Stevens Institute of Technology are related, with the imiJortant conclusion that a load of GO per cent of the ultimate strength will break a stick if left loaded (one small test piece having liee.u left loaded fifteen mouths with this result). In addition the following list of references to American work in timber physics is here inserted, with a regret that it has not been possible to include all the stray notes which may be in existence but were not accessible. Those able to add further notes are invited to aid in making this reference list complete: Abbott, Arthur V. Testing machines, their history, construction, and use. With illustrations of machines, includ- ing that at Watertown Arsenal. Van Nostrand's Magazine, 1883, vol. 30, pj). 204, 325, 382. 477. Day, Frank M., University of Pennsylvania. The microscopic examination of timber with regard to its strength. Read before American Philosophical Society, 1883. Estrada, K. D. Experiments on the strength and other properties of Cuban woods. Investigations carried on in the laboratory of the Stevens Institute. Van Nostrand's Magazine, 188S, vol. 2S), pp. 417,441. Flint, . Report of tests of Nicaraguan woods. .Journal of Franklin Institute, October, 1887, pp. 289-315. (ioodale. Prof. George L., Harvard University. Physiidogical Botany, 1885, ch.aptcrs 1, 2, 3, 5, 8, 11, and 12. Ihlseng, JIagnus C, Ph. D. On the modulus of elasticity iu some American woods, determined by vibration. Van Nostrand's Magazine, 1878, 19. On a mode of measuring the velocity of sounds in woods. Read before the National Academy of Science, 1877; published in American .Journal of Science and Arts, 1879, vol. 17. .Johnson, Thomas II. On the strength of columns. Paper read at annual convention of American Society of Civil Engineers, 1885. Transactions of the Society, vol. 15. TIMBEB PHYSICS — EARLIER WORK. 383 Kidder, F. E. Exporimrnts at Maine State College on transverse strength of southern and white pine. Van Nostrand's Magazine, 1879, vol. 22. Experiments' with yellow and white pine. Van Nostrand's Magazine, 1880, vol. 23. Experiments on the strength .and stittness of small spruce beams. Van Nostrand's Magazine, 1880, vol. 24. Influence of time on bending strength and elasticity. Journal of Fr.auklin Institute, 1882. Proceedings Institute of Civil Engineering, vol. 71. Lanza, Gaetano, professor Massachusetts Institute of Technology. Address before American Society of Mechanical Engineers, describing the 50,000 pound testing machine at VVatertown Arsenal and tests of strength of large spruce beams. Journal of Friinkliu Institute, 1883. Report of Hoston Manufacturers' Mutual Fire Insurance Company of tests made with Watertown machine on columns of pine, whitewood, and oak of dimensions used in cotton .and woolen mills. See summary and tables of sami- in Burr's Elasticity and Resistance of the Materials of Engineering, p. 480. Macdonald, Charles. Necessity of government aid in making tests of materials for structural purposes. Paper re.ad before the American Institute of Mining Engineers. Van Nostrand's Magazine, 1882, vol. 27, p. 177. Norton Prof. \V. A., Yale College. Results of experiments on the set of bars of wood, iron, and steel after a transverse set. Experiments discussed in two papers read before the National Academy ses. Philbrick. Professor, Iowa University. New practical formulas for the resistance of solid and built beams, girders, etc., with problems and designs. Van Nostrand's Magazine, 1886, vol. 35. Pike, Prof. W. A. Tests of white pine, made in the testing laboratory of the University of Minn<-sota. Van Nos- trand's Magazine. 1885, vol. 34, p. 472. Rothrock, Prof. . I. T.,Uuiver8ity of Pennsylvania. Some microscopic distinctions between good and bad timber of the same species. Read before American Philosophic Society. Smith, C. Shaler, C. E. Summary of results of 1,200 tests of full-size yellow-pine columns. See W. H. Burr's Elasticity and Resistance of the Materials of Engineering, pp. 485-490. Thurston, Prof. R. H., Cornell University. The torsional resistance of materials. Journal of Franklin Institute, 1873, vol. 65. Experiments on torsion. Van Nostrand's Magazine, July, 1873. Experiments on the strength, elasticity, ductility, etc., of materials, as shown by a new testing machine. Van Nostrand's Magazine, 1874, v(d. 10. . The relation of ultimate resistance to tension and torsion. Proceedings of Institute of Civil Engineers, vol. 7, 1878. . The strength of American timber. Experiments at Stevens Institute. Paper before A. A. A. S., 1879. Journal of Franklin Institute, vol. 78, 1879. . Effect of prolonged stress upon the strength and elasticity of pine timber. Journal of Franklin Institute. vol. 80, 1880. . Influence of time on bending strength and elasticity. Proceedings A. A. A. S., 1881. Proceedings Institute of Civil Engineers, vol. 71. Watertown Arsenal. Summary of results of tests of timber at, in Ex. Doc. No. 1, Forty-seventh Congre.ss, second session. See Burr's Elasticity and Resistance of Materials of Engineering, pp. 486 and 535. Wellington, A. M., c. e. Experiments on impregnated timber. Railroad Gazette, 1880. OEGANIZATION AND METHODS. Although in the course of the investigations many minor and some more important changes in methods became necessary, the general plan -was in the main adhered to. We consider it, therefore, desirable to restate from the same bulletin such portions as will explaiu the methods pursued. The work at the test laboratory at St. Louis, Mo., was described in full by Prof. J. B. Johnson, in charge, and the metbods in the examination of the physical properties of the test material by the writer. There are four departments necessary to carry on the work as at present organized, namely : (1) The collecting department. (2) The department of mechanical tests. 384 FORESTRY INVESTKiATIONS U. S. DEPARTMENT OF AGRICULTURE. (3) Tbe department of physical and microscopic examination of the test material. (4) The department of compilation and final discnssion of results. The region of botanical distribution of any one speoics that is to be investigated is divided into as many stations as there seem to be widely different climatic or geological differences in its habitat. In each station are selected as many sites as there seem widely different soils, elevations, exposures, or other stiiking conditions occupied by the species. An expert collector describes carefully the conditions of station and site, under instructions and on blanks appended to this report. From each site five mature trees of any one species are chosen, four of which are average representatives of the general growth, the fifth, or "check" tree, the best developed that can be found. The trees are felled and cut into logs of merchantable size, and from the butt end of each log a disk inches in height is sawed. Logs and disks are marked with numbers to indicate number of tree and number of log or disk, and their north and south sides are marked ; their height in tiie tree from the ground is noted in the record. The disks are also weighed immediately, then wrapped in oiled paper and packing paper, and sent by mail or express to the laboratory, to serve the ijurpose of physical and structuial examination. Some disks of the limbwood and of younger trees are also collected for other physical and physiological investigations, and to serve with the disks of the older trees in studying the rate of growth and otlier problems. The logs are shipped to the test laboratory, there sawed and i>repared for testing, carefully marked, and tested for strength. The fact that tests on large pieces give different values from those obtained from small pieces being fully established, a number of large sticks of each species and site will be tested full length in order to establish a ratio between the values obtained from the different sizes. Part of the material is tested green, another part when seasoned by various methods. Finally, tests which are to determine other working qualities of the various timbers, such as ada])t them to various uses, are contemplated The disks cut from each log and correspondingly marked are examined at the botanical labora- tory. An endless amount of weighings, measurings, countings, computings, microscopic examina- tions, and drawings is required here, and recording of the observed facts in such a nninner that they can be handled. Chemical investigations have also been begun in the Division of Chemistry of the Department of Agriculture, the tannic contents of the woods, their distribution through the tree and their lelation to the conditions of growth forming the first seiies of these investigations. It is evident that in these investigations, can ied on by competent observers, besides the main object of the work, much new and valuable knowledge unsought for must come to light if the investigations are carried on systematically and in the comprehensive plan laid out. Since every stick and every disk is marked in such a manner that its absolute position in the tree and almost the absolute position of the tree itself or at least its general condition and surroundings are known and recorded, this collection will be one of the must valuable working collections ever made, allo.w- iug later investigators to verify or extend the studies. This significant prophetic language also occurs in this connection, which has finally been realized by the discovery of the relation between compression and beani strength: I3y mid liy it is expected that the luimljer of tests uecessary may Ije reduced considerably, when for eacli species tlie relation of the different exhibitions ofstrengtli can be sufficiently established, and ]ierhaps a test for compres- sion alone furnish sufficient data to compute the strength in other directions. WORK AT TIIE TEST LABORATORY AT ST. LOUIS, MO. SAWING, STOniNG, AND SEASONING. On arrival of the logs in St. Louis they are sent to a sawmill and cut into sticks, as shown in fig. 103. In all cases the arrangements shown in Nos. 1 and 2 are used, except when a detailed study of the timber in all parts of the cross section of the log is intended. A few of the most perfect logs of each species are cut up into small sticks, as shown in Nos. 3 and 1. The logs tested for determining the effects of extracting the turpentine from the Southern pitch jjines were all cut into small sticks. In all cases a "small stick" is nouiiually 4 inches square, but when dressed down for testing may be as small as 3i inches si|Uiire. The "large sticks" vary from 6 by 12 to 8 by 16 inches iu cross section. All logs vary from 12 to IS feet in length. They all have a north and south diametral line, together with the numlier of the tree and of the log plainly marked on their larger or lower ends. The stenciled lines for sawing are TIMBER PHYSICS — TESTING. 385 adjusted to this north and soiith line, as sliowu in the figures. Each space is then branded liy deep dies with three 25 numbers, as, for instance, thus ; ', which signifies that this sticli was number 4, in log 2, of tree 2.5. A facsimile of 4 the stenciling is recorded in the log book, and the sticks there numbered to correspond with the numbering on the logs. After sawing, each stick can be idcntilied and its exact origin determined. These three numbers, then, become the idoutificatiou marks for all 8|>ecimeus cut from this stick, and they accompany the results of tests in all the records. The methods of sawing .shown in Nos. 2 and 4 are called "boxing the heart;" that is, all the heart portion is thrown into one small stick, which in practice may be thrown away or put into a lower grade without serious loss. In important bridge, floor, or roof timbers, the heart should always be either excluded or " boxed" in this way, since its presence leads to ebeclving and impairs the strength of the stick. .\rter sawing, the timbers are stored in the laboratory until they are tested. The "greeu tests" are made usually within two months after sawing, while the "dry tests" are made at various subsequent times. One end ((iO inches) of each small stick is tested green, and the other end reserved and te-sted after seasoning. The seasoning is hastened in some cases by means of a drying box. The temperature of the inflowing air in this drying box is kept at about 100 ' F., with suitable precaution against checking of the wood, and the air is exhausted by means of a fnn. The air is, therefore, somewhat rarefied in the box. The temperature is at all times under control. It operates when the fan is running, and this is only during working hours. The mechanical and moisture te>st are then made according to known methods. Jf / -^ ' i ■0\ Y \ v\ / \ 1 . ,y/ Fio. 10.3. — Method of sawing test logs. EXAMINATION INTO THE PHYSICAL PROPBETIES OF TEST MATERIAL. The phy.sical examination consists in ascertaining the specific weight of the dried material, and incidentally the progress and amount of shrinkage due to seasoning; the counting and measuring of the annual rings, and noting other microscopic appearances in the growth; the microscopic investigation into the relation of spring and summer wood from ring to ring; the frequency and size of medullary rays; the ^.imber of cells and thickness of their walls; and, in short, the consideration of any and all elements which may elucidate the structure and may have influence upon the properties of the test iiiece. The rate of growth and other biological facts which may lead to the finding of relation between physical appearance, conditions of growth, and mechanical properties are also studied incidentally, SHAPING AND MARKING OF THE MATERIAL. The object of this work being in part the discovery of the differences that exist in the wood, not only in trees of different species or of the same species from various localities, but even in the wood of the same tree and from the same cross section, a careful marking of each piece is necessary. The disks are split, first into a north and south piece, and each of these into smaller pieces of variable size. In one tree all pieces were made but 3 cm. thick radially, in another 4 cm., in still others 5 cm., while in some trees, especially wide-ringed oaks, the pieces were left still larger. In the conifers the outer or first piece was made to contain only sap wood. Desirable as it appeared to have each piece contain a certain number of rings, and thus to represent a fixed period of growth, it proved impracticable, at least in the very narrow-nnged disks of the pines, where sometimes the width of a ring is less than 5 mm. (0.2 inch). H. Doc. 181 25 38G FORESTRY INVESTIGATIONS U. S. DEPARTMENT OF AGRICULTURE. Some of the disks were split to a wedge sliape from center to periphery, so that each smaller piece not only represents a certain period of growth in quality, but also in quantity, thus simplify- ing the cal(;ulations for the entire piece or disk. Other pieces were left in their i)rismatic form, when to calculate the average density of the entire piece the density of eavAi smaHer piece is multiplied by the mean distance of this smaller piece from the center, and the sum of the products divided by the sum of the distances. Each piece is marked, first by the number of the tree, in Arabic; second, by the number of the disk, in Komau numbers; and if split into small pieces, each smaller piece by a letter of the alphabet, the piece at the periphery in all cases bearing the letter a. Besides the number and letters mentioned, each piece bears either the letter N or S, to indicate its orientation on the north or south side of the tree. To illustrate: 5 — vii N a means that the piece bearing the label belongs to tree 5 and disk vii comes from the north side of the tree, and is the peripheral part of this disk piece. From the collector's notes the exact position of this i>iece in the tree can readily be ascertiiined. The entire prisms sent by Ireight are left in the original form, unless used for special purposes, and are stored iu a dry room for future use. WEIfilllNG AND MEASURING. The weighing is done on an apothecary's balance, readily sensitive to 0.1 gram with a load of more than 200 grams. Dealing with pieces of iJO(» to 1,000 grams in weight, the accuracy of weigh- ing is always within 1 gram. The measuring is done by immersion iu an instrument illustrated in the following design: I'is a vessel of iron ; »S' represents one of two iron standards attached to the vessel and i)rq)ectiug Fio. 1(14. — Apitiiratiis lor ilntenniniiig apt'citid gravity above its top; B is a metal bar fastened to the cup A, which serves as guard to the cup and pre- vents it going down farther at one time than another by coming to rest on the standards S. The cup A dips down one-sixteenth to one-eighth of an inch below the edge of the knee-like spout. In working, the cup is lifted out by the handle which the bar B forms, water is poured into the vessel until it overflows through the spout, (hen the cup is set down, replacing the mobile and fickle natural water level by a constant artificial one. Now the instrument is set, the pan 1' is placed under the spout, the cup is lifted out and. held over the vessel, so that the drippings fall back into the latter, the piece of wood to be measured is put into the vessel and the cnp replaced, antl pressed down until the bar /> I'ests on the staud;irds T,!^ • . • , Fresh weight — dry weight. 1 4) Moisture in wood = rs-^, -■-,-, — Fresh weight. Tn presenting these values they are always multiidied by 100, so that the density expresses the weight of 100 cm.^ of wood; thus the shrinkage and the amount of moisture become the shrinkage and moisture jter cent. .SHRINKAOE EXPEUniF.NTS. To discover more fully the relations of weight, humidity, and shrinkage, as well as "checking" or cracking of the wood, a number of separate experiments were made. A number of the fresh specimens were weighed and measured at variable intervals until perfectly dry. Some dry pieces were placed in water and kept immersed until the maximum voluiu.e was attained. Without describing more in detail these tests and their results, it may be mentioned that in the immersed jjieces studied the liiial maximum volume differed very little, in some cases not at all, from the original volume of the wood when I'resh; and also that in a piece of white pine only 15 en;, long and weighing but 97 gs. when dry, it required a week before the swelling ceased. To determine the shrinkage in different directions a number of measurements are made in pieces ot various sizes and shapes. In most cases pins were driven into the wood to furnish a lirm metal point of contact for the caliper. A number of pieces of oak were cut in various ways to study the etlect ot size, form, and relative position of the graiu ou checking. WOOD STRUCTURK. The most time-robbing, but also the most fascinating, part of the work consists in the study ol the wood as an important tissue of a living organism; a tissue where all favorable and unfavorable changes experienced by the tree during its long lifetime find a permanent record. GKNERAL APPEARANCE. For this study all the specimens from one tree are brought together and arranged in the same order in which they occurred in the tree. This furnishes a general view of the appearance of the stem; any striking peculiarities, such as great eccentricity of growth, unusual color, abundance of resin in any part of the stem, are seen at a glance and are noted down. A table is prepared with separate columns, indicating — (1) Height ot the disk in the tree (this being furnished by the collector's notes); (2) Radius of the section ; 388 POKESTRY INVKSTIGATIONS TT. S. DEPARTMENT OF AGRICULTUEE. (3) Niimlxir of rings from perijjliery to center; (4) Number of rings in the sapwood; (5) Width of the sapwood; and (0) Kemarks on color, grain, etc. The results from each disk occupy two lines, one for the i)ieces from the north side and one for those of the south side. The radius is measured correct to one-half millimeter (0.02 iiuih), and the figures refer to the air-dry wood. To couut tlie rings, the piece is smoothed witli a sharp knife or jjlane, th(> cut being made oblique, i. e., not quite across the grain, nor yet longitudinal. Beginning at the periphery, each ring is marked with a dot of ink, and each tenth one with a line to distinguish it from the rest. After counting, the rings are measured in groui)s of ten, twenty, thirty, rarely more, and these meas- urements entered in separate subcolumns. In this way the rate of growth of the last ten, twenty, or thirty years throughout the tree is found, also that of similar periods jjrevious to the last; in short, a fairly complete history of the rate of growth of tlie tree from the time when it had reached the height of the stump to the day when felled is thus obtained. Not only do tliese rings furnish information concerning the gi-owth in thickness, but indicating the age of the tree when it had grown to the height, from which the second, third, etc., disks were taken, tlie rate of growth in height, as well as that of thickness, is deterndned, any unfavorable season of growth or any series of such seasons are found faithfully recorded in these rings, and tlie influence of such seasons, whatever their cause, both on the (|uantitj^ and on the (juality or properties of the wood, can thus be ascertained. In many cases, especially in the specimens from the longleaf pine, and from the limbs of all pines, the study of these rings is somewhat diflScult. Zojies of a centimeter and more exist where the width of the lings is such that the magnifier has to be used to distinguish them. In some cases this difficulty is increased by the fact that the last cells of one year's growth differ from the first cells of the nest year's ring oidy in form and not in the thickness of their walls, and therefore jiroduce the same color effect. Such cases frequently occur in the wood of the upper half of the disks from limbs (the limb supported horizontally and in its natural position), andoften tlie magnifier has to be reenforced by the microscope to furnish the desired information. For this puri)osc the wood is treated as in all microscopic work, being first soaked in water and then sectioned with a sharp knife or razor and examined on the usual slide in water or glycerin. The reason for beginning the counting of rings at the periphery is the same which suggested the marking of all periplieral pieces by the letter a. It is convenient, almost essential, to have, for instance, the thirty-fifth ring in Section II represent the same year's growth as thethirty-flfth ring in Section X. The width of the sapwood, the number of annual rings composing it, as well as the clearness and uniformity of the line separating tlie sapwood from the heartwood, are carefully recorded. In the columns of " remarks" any peculiarities which distinguish the particular piece of wood, such as defects of any kind, the presence of knots, abundance of resin, nature of the grain, etc., are set down. When finished, a variable number, commonly :i to (5 small pieces, fairly representing the wood of the tree, are split off, marked with the numbers of their respective disks, and set aside for the microscopic study, which is to tell us of the cell itself, the very element of structure, and of its share in all the properties of wood. The small pieces are soaked in water, cut with a sharp knife or razor, and examined in water, glycerin, or chloriodide of zinc. The relative amount of the thick-walled, dark-colored bands of summer wood, the resin ducts, the dimensions of the common tracheids and their walls, both in spring and summer wood, the medullary rays, their distribution and their elements, are the principal subjects in dealing with coniferous woods; the (luantitative distribution of tissues, or how much space is occupied by the thick-walled bast, how much by ves.sels, how much by thin- walled, pitted tracheids and parenchyma, and how much by the medullary rays; what is the relative value of each as a strength-giving element; what is the space occupied by the lumina, what by the cell walls in each of these tissues — these are among the important points in the study of the oaks. Continued sections from center to periphery, magnified 25 diameters, are employed in finding the relative amount of the summer wood ; the limits of the entire ring and that of spring and TIMBER PHYSICS STATEMENT OF RESULTS. 3«y M.W.68 v.. 57 K .■35;-. S.11% m 18 \S.6.9% lW..73Ttv.Tn~ \S. 97. llW:.7Sm.m. 245mjn\D..64 Aw. 317, . 9.4% lS2m.m. V.53 FT. W- 3«% V.Sl W337. s.s.ez R.W.Sim-m.. 191jnrm^V0 .57 fr7I63W>iGsW-33X \s.iox LH.TK.Mm.T. 175m.m. S..59 FT. ~W33Z W.'337.tiBe_ ^JSKl ,D..e4 FT. B29/liNes\s.W.'9tn.m. £W.43% D.X ir.iozlsr 90 5. 7.67.1 H.W.T 1)..36 .S-. 7.2-: jnr.73 .34 r!.67W3SZ S.6.Z% njn^.59m.Tn. ...38 FT. lis \W.40Z "•■"'■ I.S.7.3% .W..79m.m. J74/IINes \D..3e FT.W.™.„1^38% wr.777n.?n. J>..Jlf 140% S. 7.9Z R.W.gSmm. D..37 W347. ,S.9.7% liW.83m.yn. ilS^I>i«eS_ YD .39 FT. 207m.77t. "K^^^t Tr.9Jnt.7n. .43 '39~. S.9.4Z iJf.IT. 977n.iit. t97RlNes\D..4-t FT. 180m.m7\R^W-91m.m. Fig. 105 Result of physit-al examination. (Sample.) LoNULKAF I'LNE (P. jmbislris), trees. Locality : Wallace, Ala. Site: Uplaml forest, iiiiito iloiiso. Soil: Saiiiiy. White pine {P. Strubas), tree 116. Localitii : Maratlion County, Wis. Sitt:: Grown in elun.sc iiii.Ked forest. Soil: Sandy, with sandy subsoil. Legend. D. Denotes density or .spccitic gravity of the dry wood. W. Denotes percentage of water in the fresh wood, related to its weight. 5. Denotes percentjige of jhriukage in liiln drying. li. W. Denotes width of ring (average) inmiUimcters (25mm. = l inch). S. W. Denotes percentage of summer wood as related to total wood. Koman numbers refer to number of disk, placed in iio.sition of disl^. Height is given in feet from tho ground ; scale, 10 feet= 2 inches. Radius, north and south (dotted line), in millimeter.s ; scale, 10 mm. = 0.1 inch. Median line represents the pith. Right-hand numbers relate to north side, left-hand numbers to south sid''. Outer lines represent outlines of trees. 390 FORESTRY INVESTIGATIONS U. S. DEPARTMENT OF AGRICULTURE. suiumer wood are marked on paper with the aid of the camera, and thus a panorama of the entire section is brought before the eye. Tlie liistology of the wood, the resin ducts, the tracheids and medullary rays, their form and dimensions, are studied in thin sections magnified 580 diameters and even more. Any peculiarity in form or arrangement is drawn with the camera and thus graphi(;ally recorded; the dimensions are measured in the manner described for the measurement of the summer wood, or with the ocular micrometer. In measuring cell walls the entire distance between two neighboring luinina is taken as a "double wall,'' the thickness of the wall of either of the two cells being one-half of this. Tlie advantage of this way of measuring is ai)parent, since the two points to be marked are in all cases perfectly clear and no arl)itrary positions involved. The length of the cells is found in the usual way l)y se])arating the elements with Schultze's solution (nitric acid, chlorate of potassium). All results tabulated are averages of not less than ten, often of more than one hundred, measurements. In the attempt to find the quantitative relations of the different tissues, as well as the density of each tissue, various ways have been followed. In some cases drawings of magnified sections were made on good, even paper, the different parts cut out, and the paper weighed. In other cases numerous measurements and computations were resorted to. Though none of the results of these attempts can be regarded as perfectly reliable, they have done much to point out the relative importance of different constituents of the wood structure, and also the possibility and practica- bility, and even the necessity, of this line of investigation. INSTRUCTIONS AND BLANK FORMS, WJTH ILLUSTRATIVE RECORDS. Instructions fok the Collection ok Test Pieces ok Pines for Timber Investioations. A. — ohject ok work. The collector fhould undcrst.and that the ultimate object of these investigations is, if possible, to establish the relation of fjiiality of timber to the couditious under which it is grown. To accomi)li8h this object he is expected to furnish a very careful description of the conditions under which the test trees have grown, from which test pieces are taken. Care in ascertaining these and minuteness and accuracy of description are all-important in assuring proper results. It is also necessary to select and prepare the test pieces esactlj' as described and to make the records perl'ect as nearly as possible, since the history of the material is of as much importance as the determination in the laboratory. Ji. — localities for collecting. As to the locality from which test trees are to be taken, a distinction is niade into station and site. By station is to be understood a section of country (or any jilaces within that section) which is characterized in a general way by similar climatic conditions and geological formaticm. "Station," then, refers to the general Geographical situation. "Site" refers to the local conditions and surroundings within the station from which test trees are selected. For example, the drift deposits of the Gulf Coast jilain may be taken fur one station; the liuiestono country of northern Alabama for a second. But a limestone formation in West Virginia, which differs climatically, would ni^eessitate another station. Within the first station a rich, moist hummock may furnish one site, a sandy piece of upland another, and a wet savannah a third. Within the second or third station a valley might furni.sh one site, the top of a hill another, a different exposure may call for a third, a drift-capped ledge with deeper soil may wan-ant the selection of another. Choice iif stations . — For each species a special selection of stations from Avhich test pieces are to be collected is necessary. Those will be determined, in each case separately .as to number and location, from this otHce. It is proposed to cover the field of geographical distribution of a given species in such a manner as to take in stations of climatic difference and different geological horizon, neglecting, however, for the present, stations from extreme limits of distribution. Another factor which will determine choice is character of soil, as dependent upon geological formations. Stati(ms which promise a variety of sites will be preferably chosen. Choice of site. — Such sites will be chosen at each station as are usually occupied by the species at any one of the stations. If unusual sites are found occupied by the species at any one of the stations it will be determined by special correspondence whether test pieces are to be collected from it. The determination of the numlier of sites at each station must be left to the .judgment of the collector after inspection of the localities; but before determining the number of sites the reasons for their selection must be reported to this office. The sites are characterized and selected by ditl'erences of elevation, exposure, soil conditions, and forest conditions. The difference of elevation which may distinguish a site is provisionally set at .500 feet; that is, with elevation as the criterion for choice of stations the difference must be at least 500 feet. Where differences of exposure occur a site .should be chosen on each of the exposures present, keeping as much as possible at the same elevation and under other similar conditicms. Soil conditions may vary in a number of directions, in mineral composition, pliysical properties, depth, and nature of the sntisoil. For the present, only extreme differences in depth or in moisture conditions (drainage) and decided dillerence in mineral compoBition will be considered in making selection of sites. TIMBER PHYSICS 00LI.ECTIN(3 MATERIAL. 301 Forest e(in of making test pieces. (a) Mature trees. (1) Before felling the tree, blaze and mark the north side. (2) Fell tree with the saw as near the ground as practicable, avoiding the flare of the butt and making the usual kerf with the ax opposite to the saw, if possible, so as to avoid north and south side. If necessary, square off the butt end. (3) Before cutting off the butt log mark the north side on the second, third, and further log lengths. (4) Measure off and cut logs of merchantable length .and diameters, beginning from the butt, noting the length and diameters in the record. Should knots or other imperfections, externally visible, occur within 8 inches of the log mark, make the cut lower down or higher up to avoid the imperfection. (5) Continue measuring the full length of the tree and record its length. Note also distance from the ground and position on the tree (whether to the north, south, west, east) of one large sound limb. Mark its lower side and saw it of!;' close to the trunk and measure its length and record it, the limb to be utilized as described later. If the tree after felling prove unsound at the butt, it will he permissible to cut off as much or as little as necessary within the first log length. If sound timber is not found in the first log, the tree must be discarded. Only sound timber must be shipped. Any logs showing imperfections may be shortened. Be careful to note change in position of test pieces. (6) Mark butt end of each log with a large N on north side. Saw off squarely from the bottom end of each log a disk 6 inches long, and beyond the log measure cut off disks every 10 feet up to 2-inch diameter. Place eack disk ' Only men familiar witU felling and cutting tiniber should be chosen, 392 FORESTRY INVESTIGATIONS U. S. DEPARTMENT OF AGRICULTURE. on its bottom end, and after having ascertainL-d and niarlioces. (8) AVrap each piece in two sheets of heavy wrapping paper aud tie securely. (9) Mark on the newly cut bottom end of each log with a heavy pencil a north and south line, writing N on the north aud S on the south side of the log, large and distinct. Also mark centrally with an Arabic nuTubcr on each log the number of the tree in the series, and with a distinct Roman number the serial unmber of the log in the tree, counting the butt log as first. ^ Tack to the butt end of each log securely a card (centrally), on which is written name of tree, species, ioiality from which tree is taken, denoted by the letter coircspondiug to that used iu the notebook, number of tree, aud section. This card or tag is intended to insure a record of each log iu addition to the marking already made. (10) Limb wood. — Having, as before noted, selected a limb, measured and recorded its ment; also such notes as may seem desirable to complete the record and to give additioual explanations in regard to the record and suggestions respecting the work of collecting. Original records aud notes must be jjreserved, to avoid loss iu transmission by mail. TIMBER PHYSICS — COLLECTING MATEUIAL. 393 FORM OF FIELD RECORD. (Folder.) Name of crtllector: (Charles Mohr.) Kpecios: Pinus palustria. Station (denoted by capital letter): A. State: Alabama. County: Escambia. Town: Wallace. Longitude: 86' 12'. Latitude: 31° 15'. Average altitude: 75 to 100 feet. General configuration : Plain— hills— plateau — mouutainous. General trend of valleys or hills Climatic features: Subtropical; mean annual temperature, 65°; mean annual rainfall, 62 inches. Site (denoted by small letter) : a. Aspect: Level— ravine — cove— bench — slope (angle approximately). Exposure: Elevation (above average station altitude): 125 feot. Soil conditions : (1) Geological formation (if known): Southern stratified drift. (2) Mineral composition: Clay— limestone— loam— marl— sandy loam— loamy sand — sand. (3) Surface cover : Bare— grassy — mossy. Leaf cover: AJbnmlant— scanty— lacking. (4) Depth of vegetable mold (humus): Absent — moderate— plenty— or give depth in inches. (5) Grain, consistency, and admixtures: Very line— fine— medium — coarse— porous— light — loose- moderately loose — compact — binding — stones or rock, size of (6) Moisture conditions: Wet— moist— fresh— dry— arid— well drained — liable to overflow— swampy — near stream or spring or other kind of water supply.. (7) Color: Ashy-gray. (8) Depth to subsoil (if known) : .Shallow, 3 to 4 inches to 1 foot— 1 foot (o 4 feet, deep— over 4 feet, very deep^shiftiug. (9) Nature of subsoil (if ascertainable) : Red, ferruginous sandy loam ; moderately loose, or rather slightly binding; always of some degree of dampness; of great depth. Forest conditions: Mixed timber — pure — dense growth — moderately dense to open Associated species: None. Projioi'tious of these Average height: 90 feet. Undergrowth: Scanty; in the original forest often none. Conditions in tlio open: Field — pasture — lawn — clearing (how long cleared): In natural clearings untouched by tire, dense groves of second growth of the species. Nature of soil cover (if any) : Weeds — brush — sod. Station: A. (Inside of folder.) Site: a. Species: P. palustris. TiiEE No. 3. Position of tree (if :iny special point notable not ajipearing in general descriptiou of site, exceptional exposure to light or dense position, etc., protected by buildings, note on back of sheet) : In rather dense positiou. Origin of tree (if ascertainable) : Natural seedling, sprout from stump, artificial planting. DiAMETEi: breast high: 16 inches. Height to iirst limb: 53 feet. Age (annual rings on stump) : 183. Height of stump: 20 inches. Length of felled tree: 110 feet 4 iuches. Total height: 111 feet 8 inches. No. of disk. Diat-iiice from butt. Weiffht of (-■ombined disk pieces. Ecmarks. No. of log. Bistance from butt. Length of log. Diameter, butt end. I Feet. 13 19 32 47 57 67 77 87 97 Pounds. 27 20 20 18 16 14 17 14 Crown to\ichins those of nearest trees to the N. and NE. Open toward SW. I 11 Ill IV V VI VII VUI Ft. In. 8 0. 13 8 19 8 32 S 47 8 r>7 8 67 8 77 8 P(. In. 12 4 5 4 12 4 14 4 9 4 9 4 9 4 9 4 Inches. 16J 13i 121 'U II Ill IV V VI VII VIII IX X LiMBWOOD : Distance from butt: Number of disks taken: Position on trunk : Total length: Note. — As much as possible make description by underscoring terms used above. Add other descriptive terms if necessary. 394 FOKESTUY INVESTIGATIONS U. S. PEPARTRrENT OF AGKICULTURE. SAMPLE RECORDS OK TESTS. CKO.S.S BREAKING TKST. (116. Mark, 1. 1 3. 1.0Df;tli, fiO.O inches. Ileiglit, 3.74 iiichos. lireadtb, 3.75 inches. Wliito pino. Streuf^th of extreme fiber, S TV I where /:=^.-p=. 5,600 ponntls per Bqnare inch. Modulus of (dastii'.ity =l,;i2ll,()0(i iiiiiinils per si|iia.reinch. Total nisiliiuce ^3,160 inch-i)oiiu(l8. El. Kcs., 550. Resilience, per cubic inch ^4.11 inch-pounds. El. Res., 0.65. [Nninber annnal rings per inch =19.] Jul.V 18, 1891. Load. Deflection. Micrometer. Remarks. h. m. 4 24 25 26 27 28 29 31 33 35 37 40 .200 1,000 l.COO 2, 000 2, 200 2, 400 2,600 2, 8110 3,000 3,20(1 3, 300 .042 .211 .300 .454 .511 .595 .600 .8.53 1.015 1.276 1.521 0.757 0. 926 1.065 1.169 1.226 1.310 1.405 1.5C8 1. 730 1. 991 2. 238 i Maximum lo.ad. Jfeflectian, in, iruChes. TIMBER PHYSICS METHODS. 395 CROSS BREAKING TKST. f3. Longle;if piue. Mark, 3. 11. Length, 60.0 iiicUes. Height, 3.50 inches. Breadth, 3.72 inches. [Number annual rings per inch=23.] gOOO | strength of extreme fiber; where /'=o ,-,^^^10,230 pounds per squan^ inch. Modulus of elasticity =1,760,000 pounds per8i|uare inch. Total nsilience =.^,110iu< li-punnds. El. Res., 1,7H0. Resilience, per cubic inch ^6.34 inch-pounds. El. Res., 2.28. July 20,1891. Load. 2 58 I 200 3 I 1,000 1 I 1,600 2 , 2,000 3 2,400 4 2, 800 5 3, 200 6 3, UOO 7 4, 000 8 4,400 9 4, 800 13 I 5,180 Deflec- Micro- tion. meter. .042 0. S.'iS .208 1.124 .324 1.240 .401 1.320 .481 1.397 . 5.-.8 1.474 .t>40 1.556 .721 1.637 .815 1. 731 .926 1.842 1.074 1.990 1.544 2.460 Mark. Lougloaf pine: 3 3 1 "Wliilo pine; 116 1 3 Mark. Longlcaf piue: 3 3 1 "W'liitc pine; 116 1 3 Mark. Longleaf pine : 3 3 1 White piue; 116 1 3 Deflection in inches, FINAL UECOED OB' TIMBER TESTS. Percent- age of moiHture 16.8 Cross bending test. Dimensions. Length. Inckcg. 60.0 Height. Inches. 3.50 3.74 Broadtli. Inches. 3.72 3.75 Min. 15 Load. Pounds. 5,180 Deflec- tion. Inches. 1,544 1, 521 Strength per square inch. (/) Pounds. 10, 230 5,660 Modulus of elas- ticity, (fi) Pounds. 1, 760, 000 Resilience in incli- pouudspel cubic inch, (r) Crushing endwise. Crushing across grain. Dimensions. Height. Inches. 8.1 7.6 Cross section. Inches. 3.46 3.72 3.73 3.73 Crushing load. Sq. in. \ 12.87 } 13.91 Pounds. 77, 700 48, 400 Strength per square inch. Pounds. 6,040 3,480 Dimensions. Height. Inches. 3.73 Cross section. Inches. 3.47 3.93 3.72 3.93 Crushing load. Sq. inch, \ 13.63 Pounds. 10, 400 5,200 Strength per square Tension testa. Size of re- duced sec- tion. Sq.inch. 2.38 .41 Area. Sq. inch. \ 0. 976 2.53 .45 Breaking load. Pounds. 11, 400 Strength per square inch. Pounds. 11,680 '( Shearing tests. Total shearing ,nrea. Sq. inch. 4.14 3.97 1,880 j| 4.16 4.02 Bre.akiug load. Pounds. 2,280 2.580 1,700 1,600 Sliearing strength. Pounds. 551 650 400 398 p.u'nq