本文从现有的洛特卡定律参数n、c的估算方法出发,分析了以最小二乘法为基础的最优拟合概念,指出了回归分析的最优本质,及其与k—s检验最终要求的不一致性,从而提出了舍弃最小二乘法的新的参数n、c的估算方法,并给出严格的数学证明,同时用文献[1]及[6]中的数据进行验证。在上述工作的基础上,宽宏地讨论了它们的优劣。最后,在严格的意义下,指出了不应再使用旧法的缘由。 In this paper, starting from the existing methods of estimating Lotka’s law n, c, the concept of optimal fitting based on least square method is analyzed, and the optimal essence of regression analysis is pointed out. Finally, The inconsistency of the proposed method, and then proposed a new method of estimating the parameters n, c which abandons the least square method, and gives strict mathematical proofs. At the same time, it verifies with the data in [1] and [6]. On the basis of the above work, they discussed the pros and cons of the project in a broad manner. Finally, in the strict sense, pointed out the reason why the old law should not be used again.