This study was performed to find out the types of linkage of carbohydrates in wood cell walls. To study the structure of linkage of carbohydrates in wood cell walls, we have attempted to find out the method holocellulose preparation and optimum condition of enzyme hydrolysis in holocellulose, and fractionate oligosaccharide with products that hydrolized partly by acetolysis and deacetylation in holocellulose. We have achieved four results. These results as follow; 1. At first. we reacted in wood meal $NaClO_2$ 1g per lignin lg for one hour and then the same of quantity $NaClO_2$ for four hours. Through these experiments, we have developed new holocellulose preparation method which had low loss of carbohydrates and high effect of the delignification. 2. The optimum condition of enzyme hydrolysis of holocellulose which had lignin was 0.005M sodium acetate buffer (pH 5.0). We have achieved 7.2% reducing sugar through the procedure that reactioned 0.01g holocellulose putting enzyme 0.03g for 72 hours. It may be supposed that 5.5% of lignin contained in holocellulose prevented enzyme contaction from holocellulose and so this lignin has resulted in the low efficiency of enzyme hydrolysis. 3. We did not fractionated from oligosaccharides which were preparated by the method of acetolysis and deacetylation in holocellulose. The reason is that holocellulose having a lot of lignin prevented prefectly partial hydrolysis from the method of acetolysis and deacetylation. 4. We attempted analysis of six standard substances through HPLC apparatus having sugar pak 1 column which we have changed flow rate and the column temperature variably. These six standard substances were D-glucose, D-mannose, D-xylose, D-galactose and L-rhamnose, L-arabinose, But sugar pak 1 column was not fitted analysis of four substances because D-galactose, D-mannose, D-xylose, L-rhamnose were agreement with elution time. And so, we could not analize four standard substances with sugar pak 1 column.