I. The method of interserial differences can be applied to the different (territorial) series of ethnical groups, if these series show the difference of MM. The best results can be drawn if there are neighbouring series. This method can be considered as an analytical method, but the results of application must be always verified by other analytical methods, because this method is based on MM, which sometimes confuse the real anthropological components of the ethnical groups under discussion.
II. The method of correlation must be always applied. If the coefficient has no numerical value or very insignificant one the analysis of the series can be done by superimposing of the fields of correlation of the same characters, but belonging to different groups or series. The insignificant coefficients resulting from the calculation of the amalgamated series composed of many anthropological types have no importance as such, but such coefficients cannot be considered as absolutely meaningless, if other series taken in comparison show significant coefficients or opposite signs. Thus this method may be applied comparatively to the analysis of the territorial and ethnical Series. For instance the differentiation of the cephalic index and similarity of the stature of two types will show no correlation, but this result can be controlled by the correlation of the stature and cephalic index with other measurements and indices separately. Then, perhaps, there will be discovered some more complicated amalgamations in which may be distinguished two types of the same stature and different cephalic index, and, on the other hand, two types of different stature and similar cephalic index. At the same time all these types may be distinguishable by other characteristics, as for example nasal and frontal indices or colour of the skin and hair, and so on, which characterize four different types.
Note on Section II,
It might be that in the Anthropology the application of this method cannot give at all the high numerical meaning of coefficients, except the cases of the anatomical (constructive) correlation or very simple amalgamations of the types, as for example, one type of very high stature with very high cephalic index and other type of very low stature and very low cephalic index.
III. The calculation of the coefficients of correlation and especially the putting of the individuals in the field of correlation must be done for several pairs of measurements and indices, because the anthropological types in the state of amalgamation can sometimes produce very insignificant differentiation of some characters and very significant differentiation of other ones.