Three kinds of plant materials, mills vetch, azolla and rice straw were mixed with a soil,of negligible organic matter content and allowed to decompose in field under upland and waterlogged condition respectively. Samples of each plant material were talen at intervals over a period of one year and were partitioned into light and heavy fractions, observations made on the soil fractions obtained by this procedure included the determination of C, N content, sugar content and fractional humus composition. Results obtained are as follows: 1. Although the difference in % of C remained of plant materials with different chemical composition decomposing under the same environments or of the same plant materials decomposing under different environments was partly due to the difference in the amount of undecomposed or/and partially decomposed plant material, it was found that plant materials differed from each other in the amount of humus they formed after 1 year decomposition according to their chemical composition and the environmental conditions under which they decomposed. 2. The composition of the newly formed humus also varied with both the kind of plant material they derived from and the environmental conditions under which they decomposed. The sugar content of humus followed the order of rice straw>milk vetch>azolla which is also held for the original plant materials. The C/N ratio of the newly formed humus was much lower than that of the native soil humus, with that derived from rice straw the largest being 7.0, and that derived from milk vetch the smallest, being 6.5. The HA/FA of humus derived from rice straw was the highest, and that from azolla was the lowest, while the optical density (at 465 μm) of HA followed the reverse order. No regular difference in fractional humus composition between humus. formed under waterlogged condition and that under upland conditions was found.