本文简要叙述了在矿物工业中很需要取得处理物料中有关各种矿物的比例、大小及其相互关系等准确数据。要想使它们具有最大价值,这些数据的收集就必须迅速而准确。用传统的手工方法测定这些矿物参数太慢而且带主观性,适应不了现代工业的要求。本文叙述了如何用自动测定装置来补充手工方法所测得的结果。详细描述了适当的自动测定系统所要求的设计特征以及满足许多这些要求的方法。现有的自动测定装置只是根据矿物的光学性质来鉴别矿相,但光学性质已被证明常常不适于用来进行矿物的自动鉴别。本文指出利用一个改进了的电子探针X射线显微分析仪的意义,这种仪器能给出许多同时发生的信号,综合利用这些信号,就能做出出色的矿物鉴别。详述了把显微分析仪改成用计算机控制的“试样平面”测定装置的方法。计算机能使这种装置在工作时具有很大灵活性,并能部分地评价所收集到的数据。结果是由一些能给出有关矿物大小、形状、比例和位置等资料的线性扫描而得出的。除非通过适当的立体操作将这些结果同相应的三维数值联系起来,否则它们就没有多大用处。描述了将矿物的比例、大小和形状等线性数据改变成体积数值的立体变换,并说明在全面利用立体关系的线性数据以前,还需进一步做工作。介绍了用计算机控制的显微分析仪测定连生矿物颗粒的粒度分布和粒状物料的形状以及确定矿物痕量等的实例。指出对连生矿物的立体学还需要做更多的工作,并指明仪器将来发展的方向。 This article briefly describes the exact data on the proportions, sizes and their interrelationships of various minerals in the treated material that are very much needed in the mineral industry. To maximize their value, these data must be collected quickly and accurately. The determination of these mineral parameters by traditional manual methods is too slow and subjective to meet the requirements of modern industry. This article describes how to use automated measuring devices to supplement the results of manual methods. Described in detail the design features required for a suitable automated measurement system and methods to satisfy many of these requirements. Existing automated measuring devices only identify the mineral phase based on the optical properties of the mineral, but optical properties have proven to be often unsuitable for automatic identification of minerals. This article points out the significance of using an improved electron probe X-ray microanalyzer that provides many simultaneous signals that can be combined to make excellent mineral identification. The method of changing the microanalyzer to a computer-controlled “sample plane” measuring apparatus is described in detail. The computer gives the device great flexibility in its work and can partially evaluate the data it collects. The result is a number of linear scans that give information about the size, shape, proportion, and location of the mineral. Unless associated with appropriate three-dimensional values ​​by appropriate stereoscopic manipulations, they are of little use. Described a three-dimensional transformation that changed linear data such as proportions, sizes, and shapes of minerals into volumetric values ​​and showed that further work was needed before full use of linear data for the three-dimensional relationship was made. An example of the determination of the particle size distribution of the continuous mineral particles and the shape of the granular material by computerized microscopic analyzer, as well as the determination of the trace of the mineral, is presented. It points out that more work needs to be done on the stereochemistry of contingent minerals and the direction of the future development of the instruments.