In the case of testing decay durability of wood at the laboratory, it is very important to adjust the test condition in which the wood-destroying fungi can readily grow. Above all, the moisture content of wood is most important. In the present instance, the variation of the moisture content and the weight loss of wood blocks in the decay durability test were investigated in accordance with the Japanese Industrial Standard (JIS) original bill.
   The wood used in this test were the sapwood (2×2×2 cm) of Beech (Fagus crenata), Sugi (Cryptomeria japonica) and Akamatsu (Pinus densiflora); the fungi were Poria vaporaria and Coriolus versicolor, and the medium was Beech sawdust (10 - 60 mesh) to which was added 1% glucose and 0.2% peptone and which was mixed with distilled water that was about half as much again of sawdust in weight.
   After a week followed the inoculation of fungi: two test pieces were placed perpendicular on the surface of the culture medium in each culture vessel. After each 10, 20, 30, 45, 60, 90, 120, 150 and 180 days, four test pieces, tested to each wood species and of test fungi, were taken out from the culture vessels and their moisture content and weight losses examined.
   The moisture content of wood increased suddenly about 10 days after fungal attack, but after that no particular change could be observed. So it seems that the moisture content had no influence on the weight loss (See. Fig. 2, 3 and 6).
   The rate of weight loss was little at the first short time periods, then suddenly it became faster, and after that it slowly decreased. As for Poria vaporaria, the maximum of the rate was observed from 20 to 30 days after fungal attack, and the same in the case of Coriolus versicolor on Beech, but on Akamatsu and Sugi, it was observed about two months after (See. Fig. 4 and 5).
   As mentioned above, in the case of comparing the decay durability of many species of wood with the one of Beech wood, which is used as a standard wood in the decay durability test of JIS, it seems that about two months is suitable for the duration of the test.

   In order to obtain fundamental knowledge on the properties of charcoal produced from Quercus, Fagus, and other deciduous trees which are important species in Japan, the reactivity of charcoal to carbon dioxide at 950°C by the flow process in comparison with gas-oven coke, was investigated.
   Simultaneously, the aerial oxidation of the above samples was carried out with the quartz thermo-spring balance in higher temperature.
   The results are summarized as follows:
   (1) The reactivity of charcoal is very high, and after deashing by repeated washing with hot water, it drops but yet remains much higher than that of gas-oven coke.
   (2) Relationship between the reactivity and industrial analytical values of charcoal was determined, and high value of reactivity was obtained from deciduous char which contains much volatile matter, low moisture and fixed carbon. On the contrary, low value of reactivity was obtained from Quercus white charcoal, carbonized above 1000°C, which contains low volatile matter and high moisture and fixed carbon.
   (3) The activated energy obtained at the beginning of weight loss of aerial oxidation indicated a range between 4.2∼10.7 kcal/mol. The relationship between this energy and reactivity to carbon dioxide was represented in a straight line.

   The aim of this study is to establish the coarse sand-block testing method for rapid evaluation of preserving effect of wood treated with preservative in the laboratory. This method used is coarse sand in place of soil as culture medium.
   As a first step, we studied the condition of sand culture medium for methodology of coarse sand-block testing.
   The discussion on the moisture content, nutriment, size of sand grain, disposition of test pieces, term and temperature of incubation and putting time of test pieces in the sand culture medium was tried. The results of weight loss rate and compressive strength obtained was analysed by statistics.
   According to the results, we can recommend the sand culture medium of 100∼120ml water content, 3∼4.5 g glucose, 0.2∼0.3 g peptone and 10∼40 mesh sand grain for the testing method of evaluating wood preservative.

   The composition analyses of carbohydrates of decayed pine wood revealed that the difference of decomposed glucan between brown rotted wood and white rotted wood, in the same weight loss, was not remarkable in percentages. However, it has been described in the first report of this series that the α-cellulose of wood was very rapidly reduced during decay caused by brown rotting fungus.
   Therefore, the author presumed that both the brown rotting fungus (Poria vaporaria) and the white rotting fungus (Peniophora gigantea) utilized the glucan of the samples in approximately the same degree, but the brown rotting fungus produced rapidly low molecular fractions of carbohydrates in the process of decomposition of wood.
   The hemicellulose fractions of decayed pine wood, including mannan, xylan, araban, etc., were decomposed faster than glucan (including cellulose).
   In the white rotted beech wood caused by Coriolus versicolor, the glucan and the xylan, which were major components of carbohydrates, decreased uniformly during decay.

   End-grain heartwood wafers, 1 × 1 × 1/8" in size, were prepared from six tropical woods. Materials consisted of two groups of wafers; one was the extractive-free (F) which was subjected to successive extractions with ether, methanol, 0.01N NaOH and 0.01N HCl and the other was the unextracted original (A). Wafers of both groups were exposed to attack by Polyporus versicolor and Poria monticola for 90-days in a room maintained at 24°C. Two decay-test methods, the agar-glass rod-block method, in which a 3mm. diameter bent-glass rod was used as the bed of the wafer specimen, and the agar-feeder block-block method, in which the specimen rests on the feeder-block, were employed. In the decay test, six wood species, two fungal species, one test period, two treatment groups, two test methods and four replicates were used, iis, 6 × 2 × 1 × 2 × 2 × 4 = 192 specimens, each in its individual upright square bottle conta Thatning 25 cc. malt-agar on the bottom, were used.
   The degree of decay was measured by weight loss, on the basis of moisture-free untreated wood weight before subjecting to decay. Weight-loss values for each specimen are shown in Table 1. The effects of main factors and their interactions are statistically studied in the form of the analysis of variance shown in Table 2.
   The conclusion to be drawn is, that among the noticeable results, special emphasis should be laid on the interaction between the fungal species and the test method (M×F). The effect of this interaction is highly significant, although the effect of the test method (M) is not significant in the result of lying concealed behind the average of the two oppositeuresults. That is to say, each fungus has the optimum condition for its own activities; thus by differing the test method, significantly different weight Losses might be obtained with the same fungus. One should bear this point in mind in the laboratory decay tests and in the case of revising the standard method for the decay tests.

   Kinetic parameters such as reaction order n, energy of activation E and frequency factor A for the pyrolysis of xylan under vacuum were determined by means of dynamic thermogravimetric analysis (CHATTERJEE method) and isothermal analysis (CHATTERJEE-CONRAD method).
   The result obtained by CHATTERJEE method indicated that four reaction stages were involved in the range from 200° to 300°C. The initial stage accompanied by weight loss up to 10% was controlled by zero-order kinetics and E was 39.9 kcal/mol. The subsequent first-order reaction was divided into three stages. In the first stage, E was 29.6 kcal/mol, in the second one, E = 17.7 kcal/mol and in the last, E = 43. 6 kcal/mol.
   The investigation by isothermal method (204°, 216° and 221°C) showed the existence of two first-order regions. The values of E were 25. 7 kcal/mol and 21.9 kcal/mol, respectively. Further, another zero-order step was revealed in the range from 200°C to the isothermal temperature in which about 15% of total weight was lost, and yielded E = 45.6 kcal/mol by plotting the temperature dependency of the rate of weight loss.
   Assumption of the pyrolytic reaction mechanisms of xylan have been proposed through these results.

   In a previous paper, it was found that the carbonization (heat treatment at 400°C) of HINOKI wood had an influence on its sorptive capacity and that, by its treatment, lignin and crystalline region rapidly degraded within 2 - 3 hr. In this paper, in order to find some factors of sorption hysteresis, the relation between mechanical energy loss (tanδ) and hysteresis ratio of HINOKI sapwood (Chameacyparis obtusa ENDL.) was discussed.
   The change of hysteresis ratio with carbonization had a contrary tendency to the change of moisture content in adsorption process that showed a minimum value at about 20% weight loss. And the tendency of hysteresis ratio change was similar to that of mechanical energy loss change calculated from the vibrating reed method and from stress-strain hysteresis.
   In terms of energy loss, it would seem that sorption hysteresis and mechanical energy loss in anisotropic substance are the same phenomenon. Therefore, it could be said that structual properties of wood substance, especially its Theological behavior, is a cause of sorption hysteresis.

   Each timber has a characteristic nature of termite-resistance. In the present paper, the relative values of antitermitic activities of various species of Okinawan woods against the subterranean termite, Coptotermes formosanus (SHIRAKI), were estimated. The following two methods were adopted to calculate the antitermitic values quantitatively.
   1) Each test timber was exposed to 330 termites (300 workers and 30 soldiers) at room temperature and relative humidity of ca. 98%. After feeding for 40 days, the percentage of weight loss of timber due to the termite attack was determined.
   2) The antitermitic values were also measured from the difference between initial and final weight of termites.
   Wood materials for the termite tests were prepared and examined in the forms of block, shaving, and meal respectively.
   The contribution of extractives was different among the wood species; especially those of Inumaki (Podocarpus macrophullus LAMB), Sugi (Cryptomeria japonica D. DON), and Sendan (Melia Azedarach LINN) were remarkable.

   The thermal degradation of xylan was studied with both the dynamic heating and isothermal heating(170°, 180° and 190°C) in vacuo. The pyrolysis mechanism was investigated by determining the changes in intrinsic viscosities of the degradation products.
   The results obtained showed that the chain scission in early stage of pyrolysis occurred at random and obeyed first-order kinetics. The activation energy was 30.7 kcal/mole for the dynamic heating and 32.6 kcal/mole for the isothermal one.
   Moreover, considering the weight loss and the number of chain scissions per chain in the region of 160° to 230°C in the dynamic pyrolysis, the following mechanism of xylan pyrolysis was suggested:
   1) the initiation reaction of random chain scission,
   2) the splitting off of monomer units (propagation reaction),
   3) the stabilization of chain end (termination reaction).
   From the change of molecular weight distribution shown by gel permeation chromatography (GPC), it was revealed that the thermal degradation of xylan was caused by random chain scission.

   A heterogeneous hydrolysis of cotton cellulose was investigated at 60, 70, 80°C in 1, 2.5, 4N HCl solutions. The observed rate based on weight loss of the residue was expressed as the following equation, adopted from Philipp, Nelson and Ziffle: y = A_{t = 0}e^-k't + C_t = 0e^-kt, where, y is the fraction of the residue, A_{t = 0} and C_t = 0 are the apparent amount of the amorphous and crystalline region respectively, and k and k ' are the corresponding rate constants of hydrolysis. These constants were statistically estimated. The values obtained are comparable to those of Daruwalla and Shet. The so-called "recrystallization" was assumed to take place during hydrolysis; i. e. while a rapid hydrolysis in the amorphous region and a slow hydrolysis in the crystalline region proceed and the cellulose chains in the amorphous region cleave randomly, a portion of cleaved chains rearranges itself to form crystalline region. On the basis of this assumption, the following relationship between A_{t = 0} and k ' can be deduced, 1/A_{t = 0} = (1/a₀)(1 + γ/k '), where, k ' refers to the probability of cleavage of the chain in the amorphous region. The constant α₀ means an intrinsic value of amorphous region in the sample, and γ is a fraction of the rate constant for rearrangement of the mobile segment produced by the chain cleavage. This equation is in keeping with the results within the range of experimental error. The estimated values for α₀ was 0.1 and γ was 10^-4 - 10^-5sec^-1. Another set of estimation for γ calculated from Daruwalla and Shet's data tends to give a somewhat lower value, resulting an activation energy 5 - 25kcal/mole.

   To find the change of lignin structure with alkaline treatment, we hydrolyzed MWL of Picea jezoensis CARR. in 30%-NaOH at 100°C, 150°C and 200°C for 0.5 hour and determined the functional groups of these degradation products and analyzed them by UV, IR and NMR spectroscopy.
   Phenolic hydroxyl groups of alkaline hydrolysis products determined by Js-method and total hydroxyl groups of untreated MWL determined by Hillis-method were in good agreement with those determined by NMR-method. Therefore, NMR spectroscopy can be used for determination of phenolic and total hydroxyl groups in MWL.
   Phenolic hydroxyl groups increased remarkably at 100°C-150°C, but not at 150°C-200°C. This indicates that cleavage of ether linkage finishes at a temperature between 100°C and 150°C. The higher the reaction temperature, the less the aliphatic hydroxyl and methoxyl groups. Aliphatic hydroxyl and methoxyl groups per C₉ unit decreased by about 0.30 and 0.12, respectively.
   Hydrolysis at 150°C results in weight loss of about 15%, loss of about 0.7 carbon atom of side chain per monomer unit, and increase in condensed type unit in lignin from 36% to 63%. These results indicate that cleavage of ether linkage and side chain and condensation of aromatic ring occur predominantly at 100°C-150°C.
   We conclude that reaction mechanism suggested by model compounds with alkaline hydrolysis can be applied to alkaline hydrolysis of high molecular lignin.

   The thermal degradation of the arabinogalactan and O-acetyl- and deacetylated-galactoglucomannans which had been elaborately prepared was investigated using the techniques of thermogravimetric analysis (TGA) and differential thermal analysis (DTA) in air, flowing nitrogen and vacuum at a heating rate of 2.5°C/min. Values of kinetic parameters were calculated on the basis of TGA under vacuum.
   These results showed that all the three samples pyrolyzed in three reaction stages, and that the O-acetyl-galactoglucomannan was the most thermally unstable. On the pyrolysis of the arabinogalactan and deacetylated galactoglucomannan, a first-order reaction zone with the low activation energy (6-8 kcal/mole) was found before the region in which the rapid weight loss took place. In the latter region, the degradation was followed by another first-order with the activation energy of 28.0 kcal/mole for the deacetylated galactoglucomannan, and of 34.8 kcal/mole for the arabinogalactan, respectively. For the O-acetyl-galactoglucomannan, the first-order reaction zone with the high activation energy (38.2 kcal/mole) was found which might possibly be caused by the hydrolysis of the O-acetyl groups. The O-acetyl groups had a profound effect on the thermal degradation of the galactoglucomannan in the atmospheres used except the nitrogen.
   The corresoonding DTA curves seemed to reflect the distinct reaction zones as described above.

   Thermal degradations of 14 species of woods have been investigated using the technique of thermogravimetric analysis (TGA) between room temperature and 600°C.
   TGA curves and differential thermogravimetric curves (DTG) were obtained at 10°C/min in flowing nitrogen and air atmospheres.
   Initial significant weight loss began at 193-225°C in nitrogen and at 185-205°C in air.
   DTG curves of wood of broad-leaved tree showed the tendency of constancy or decrease of the rate of weight loss in temperature range 283-300°C in air and 295-315°C in nitrogen, but these of wood of coniferous tree and red lauan showed no such phenomena.
   The most active pyrolyses occurred at 350-370°C in nitrogen and at 320-330°C in air.
   Assuming the first-order reaction for thermal degradation, the activation energies were calculated from the data obtained by dynamic TGA.
   Thermal degradation took place by four reaction stages in temperature range 250-380°C in nitrogen and three reaction stages in 240-330°C in air except for some woods studied.
   The activation energies were as follows: in nitrogen, 16-30kcal/mole at 250-285°C, 3-14 kcal/mole (wood of broad-leaved tree) and 16-30kcal/mole (wood of coniferous tree) at 285-320°C, 20-44 kcal/mole at 320-340°C, 42-89 kcal/mole at 340-370°C, in air, 22-38 kcal/mole at 240-270°C, 11-19 kcal/mole (wood of broad-leaved tree) and 22-30 kcal/mole (wood of coniferous tree) at 270-300°C, 54-82 kcal/mole at 300-330°C respectively.

   Structural plywoods composed of face veneers of red seraya and core veneers of white seraya, were treated with 2 % Wolman salt solution by pressure, dipping and coating processes. Untreated and treated specimens were subjected to decay durability test by soil-block method, with or without preliminary weathering, against two fungi-Coriolus versicolor and Tyromyces palustris, estimating resistance to decay on weight and compression strength.
   The specimens of face veneers of fiber direction at 0° and 90°, treated with pressure process or with dipping process without weathering, showed satisfactory durability to decay.
   Concerning the kinds of fungi, in the correlation of the reduction of weight with the residual compression strength, the reduction of residual compression strength brought about with Tyromyces palustris was more remarkable than that brought about with Coriolus versicolor in the case of the equal reduction of weight over 6-8 % of weight loss.
   In the case of untreated original plywood, the decay of core veneers progressed more rapidly than that of face veneers: it is possible to attribute it to the difference of decay durability of both wood species of veneers.

   The viscosity and the weight of linter cellulose isothermally heated in vacuo were measured to study the early stage of the pyrolysis.
   The number average degree of polymerization (DP) of the cellulose, which was rapidly lowered at first, attained to the leveling off value of about 100. The decrease in the DP is explained by means of a chain mechanism of random-scission initiation and grafting or cross-linking termination. The activation energies and the pre-exponential factors for the random scission and the termination are 38.0 kcal/mole and 1.4×10¹⁰ sec^-1, and 31.4 kcal/mole and 2.9×10⁹ sec^-1, respectively. Those for the propagation, which is supposed to determine the weight loss by producing volatile levoglucosan, are estimated to be 31.5 kcal/mole and 8.9×10⁹ sec^-1 on the basis of the above values and the Arrhenius parameters for the weight loss.
    According to the proposed mechanism, the leveling off of the DP means a stationary state between the initiation and the termination. However, there is another possibility that it corresponds to the crystallite length.

   Methyl methacrylate (MMA) was grafted onto alkali-swollen cellulose (A-Cellulose) by the ceric ion method, and the effects of swelling on the degrees of reaction and the characteristics and properties of graft copolymer were studied in comparison with those of beating. Forty per cent of ethylenediamine (EDA) ethanol solution (EDA(E)), 80% EDA(E), 80% EDA aqueous solution (EDA(A)) and 20% NaOH aqueous solution were used as swelling agents. Hydrolyzable region of A-Cellulose was not superior to that of beaten celluloses (B-Cellulose). Alkali treatment, however, resulted in the increase in all of the degrees of reaction, especially in grafting efficiency, in contrast with beating. Alkali treatment and beating showed the same results in the increase in weight average molecular weight of grafted polymethyl methacrylate (PMMA) and the decrease in number of grafted chains. Weight loss and hydrolyzed cellulose of graft copolymer of A-Cellulose (A-G-Cellulose) by acid hydrolysis increased markedly except those treated with 40% EDA(E), but the extraction of grafted PMMA accompanied by hydrolysis was difficult in comparison with that of graft copolymer of B-Cellulose (B-G-Cellulose). Moisture regain of A-G-Cellulose and B-G-Cellulose increased with the increase in hydrolyzable region, and the treatment with 20% NaOH resulted in marked effect. These results suggest that interfibrillar and intramicellar swellings with alkalis were retained to some extent in A-G-Cellulose with packed PMMA, and that acid reactivity and moisture regain of A-G-Cellulose increased in comparison with B-G-Cellulose. Such effect of treatment increased in the following order; 40% EDA(E) < beating < 80% EDA(A) < 20% NaOH.

   Linter cellulose was treated with two kinds of fire retardants for cellulosic materials, diammonium hydrogen phosphate and sodium tetraborate decahydrate, in order to carry out studies on the effects of the salts on the pyrolysis of the cellulose. The degree of polymerization (DP) of the treated samples which were heated in vacuo was measured by viscosimetry. Furthermore, the weights of the untreated and treated samples were continuously measured with a thermobalance while being isothermally heated in vacuo.
   The change in the DP of the borate-treated cellulose with the pyrolysis can be explained by the same chain mechanism with random-scission initiation and grafting termination as previously shown for the untreated cellulose. The observed leveling off DP was higher than the untreated one and varied with temperature, being shown to be ascribable to a stationary state between the reactions. The weight loss with the pyrolysis was irregular at the early stage of the heating but expressed as a first-order reaction at the late stage. The activation energies of the initiation, the termination, and the weight loss are estimated to be 29.6, 20.5, and 49.9 kcal/mole, respectively, and the corresponding pre-exponential factors are 1.6 × 10⁷, 2.5 × 10⁵, and 2.0 × 10¹⁶ sec^-1.
   For the phosphate treated cellulose, however, the leveling off DP was not found because of the insoluble sample which was long heated, so the chain mechanism was inapplicable. The weight loss at the late stage may be expressed as an apparent first-order reaction. The obtained activation energies of the random scission and the weight loss are 24.8 and 40.7 kcal/mole, and the pre-exponential factors of these reactions are 4.6 × ⁵ and 6.9 × 10¹³ sec^-1, respectively.
   For the untreated cellulose, the activation energy and the pre-exponential factor of the weight loss as an apparent first-order reaction at the late stage are determined to be 44.7 kcal/mole and 3.9 × 10¹³ sec^-1.

   The diffusion controlled torsional stress relaxation for untreated Hinoki and Hinoki treated with DMSO and the DEA-SO₂-DMSO mixture was investigated and the following results were obtained.
   1) Two relaxation processes were observed in wet and DMSO swollen conditions. The values of activation energy E_act in these processes were 23.7 kcal/mole (Physical Relaxation I) and 21.6 kcal/mole (Physical Relaxation II), respectively. It was considered that physical relaxation I would result from a molecular motion in the amorphous region of both cellulose and hemicellulose, while physical relaxation II from that of lignin. On the other hand, in the diffusion controlled process of Hinoki which was initially in dry condition, these E_act values decreased to about one-half.
   2) A relaxation process (Physical Relaxation D) occurred in the diffusion controlled relaxation of Hinoki which was initially in dry condition and its E_act value was 12.4 kcal/mole. It was suggested that it would be due to changes in the microstructure such as inter- and intra-molecular networks. In the diffusion controlled relaxation, the value of chemical relaxation of Hinoki which was initially in dry condition was almost equal to those of physical relaxation I, II and D.
   3) The E_act value of the chemical relaxation process in sulfuric acid solution for Hinoki which was initially in wet condition was about 1/2-1/3 of that obtained from the measurements of viscosity and weight loss under the homogeneous reaction. On the other hand, those of Hinoki which was initially in DMSO swollen condition and treated with the DEA-SO₂-DMSO mixture were 28.5 kcal/mole and 34.7 kcal/mole, respectively, and these values were almost equivalent to those obtained under the homogeneous chemical reaction. In this connection, it may be suggested that the values of 28.5 kcal/mole and 34.7 kcal/mole correspond to the activation energies due to the scission of a glucosidic bond in amorphous region and in crystalline region of cellulose treated with DEA-SO₂, respectively.

   Linter cellulose was treated with ammonium bromide and sodium chloride for the purpose of making researches on effects of the salts on the pyrolysis of the cellulose. Following previous study, the degree of polymerization (DP) and the weight were measured during the heatings in vacuo.
   The DPs of the treated samples were rapidly reduced at first, but eventually levelled off. The changes in the DP of both the samples at the initial stage of the pyrolysis are supposed to be caused by random scission of the polymers. Those at the late stage may be explained by a chain mechanism with random-scission initiation and with grafting termination. The obtained activation energies of the random scission and the termination for the bromide-treated sample are 26.8 and 23.0 kcal/mole, respectively. The corresponding pre-exponential factors are 7.4×10⁸ and 1.7×10⁹ sec^-1. On the other hand, the determined activation energies of the random scission and the termination for the chloride-treated sample are 32.6 and 19.3 kcal/mole, and their pre-exponential factors are 9.3×10⁷ and 9.9×10⁴ sec^-1, respectively.
   The weight losses of the bromide-treated sample with the heating are too complicated to yield the kinetic parameters. Those of the chloride-treated sample are shown to be an apparent first-order reaction at the late stage. The activation energy and the pre-exponential factor of the reaction are determined to be 44.3 kcal/mole and 3.5×10¹⁴ sec^-1.

   Plywood samples were prepared in glues into which 1 of 7 kinds of fungicides was incorporated. A threshold amount of preservative based on dry wood weight was used to prevent decay (Table 1). Bond quality was tested on wet and dry samples (Table 2). The efficiencies of wood preservation evaluated by the following three methods are as follows:
   1) Percentage weight loss in decay test
   Na-PCP, Na-DNOC, NCH-K, NaF and TBTO incorporation made plywood resistant to Tyromyces palustris and NaF was the most effectiveness of 7 fungicides against Coriolus versicolor (Table 3).
   2) Area of wood surface covered by mycelia
   NCH-K and TBTO incorporation made plywood resistant to T. palustris but only NaF was effective against C. versicolor (Figure 1).
   3) Change of residual ratio of compression strength during decay test
   NCH-K, NaF and TBTO incorporation made plywood resistant to T. palustris but no fungicides were effective against C. versicolor (Table 3).
   As a result of the above evaluations, NaF was recommended to be incorporated in urea-melamine glue for preservation of plywood.