This paper analyzes the spatial distribution of two kinds of national wetland parks and their differences and similarities by applying GIS and the methods of nearest neighbor index, Gini coefficient and kernel density estimation. Their influencing factors from the aspects of natural geography, social economy, and cultural environment are discussed with the use of SPSS and geographic analysis methods. Results show that the different development of two kinds of national wetland parks is closely related to different decisions of competent authorities and the attention of the government. They both have the agglomerate trend across the nation in spatial distribution, imbalanced status of state-level wetland parks distribution, but their distribution densities are different. South Jiangsu-North Zhejiang, East Hubei-North Hunan and Southwest Shandong are the three highdensity national wetland park distribution centers, while South Jiangsu and Northeast Shandong form high-density urban wetland park distributions. Finally, this paper discusses the factors that cause the situation in three aspects which include natural conditions, social economic conditions and cultural environment.

Two novel low power pass transistor based XOR-XNOR circuits are proposed, UPPL (Unsymmetrical Push Pull Pass Transistor Logic) and CPPL (Complementary Push Pull Pass Transistor Logic). They both input single rail signals and output dual rail signals, which can get XOR and XNOR signals simultaneously. The output signals are full swing voltage. Hspice simulation under 0.18μm technology 1.8V voltage showed improvement on speed and power-delay product compared with some other circuits. Compared with the latest circuits, which was proposed by Mohamed Elgamel in 2003, the UPPL and CPPL circuits have 61.0% and 58.4% decreases on power delay product respectively without load. And with fanout three, they have 25.3% and 45.3% decreases respectively.

Let{X_n, n≥1} be a sequence of i.i.d. random variables with common d.f.F(x). X_{1, n}≤…≤X_{n, n}}are its order statistics. Qis the quantile function of F. It is shown that for any underlying distribution function F, the LIL for the heavily trimmed sums remains true, as long as λ and 1-λare continuity points of Q and σ(λ)>0. Moreover, a strong approximation result is obtained in this case.