The author previously reported in this journal (vol. 66, no. 4) on the method of sex diagnosis of the Japanese long bones by means of discriminant function. The present paper gives a method of sexing of the Japanese skulls and scapulae by the same way.
The materials were modern Japanese skulls of 105 (64 males and 41 females) and the scapulae of 132 individuals (76 males and 56 females), between the ages of 20 to 60 years. All of them belong to the Department of Anatomy and the Department of Legal Medicine, University of Tokyo School of Medicine. In case the skulls and the mandibles were combined, 100 individuals (60 males and 40 females) were used because the mandibles had been missing in 5 individuals.
Following measurements were made on each bone (X₁, X₂, ..are symbols used in calculations and tables in the text ; numbers in parentheses by ARTIN-SALLER, 57) : Skull: X₁. maximum length (1), X₂. maximum breadth (8), X₃. basi-bregmatic height (17), X₄. bizygomatic breadth (45), X₅. upper facial height (48), X₆. bigonial width (66), X₇. symphyseal height (69), X₈. condyloid height (70), X₉. breadth of mandibular ramus (71).
Scapula: XI .anatomical breadth (1), X₂. anatomical length (2), X₃. projective length of spine (7), X₄. length of cavitas glenoidalis (12), X₅i breadth of cavitas glenoidalis (13). Table 1 gives the sample numbers, means, sum of squares of deviations and mean squares for each measurement. Discriminant functions for the skulls and scapulae are shown in tables 2 and 3, respectively. Supposing the case in which a part of the bone would be broken, several kinds of functions were calculated for both of the skull and the scapula.
The practical method is as follows: At first, make required measurements for the bone to be diagnosed its sex, secondly compute Y, discriminant function, from the acquired measurement values, and finally compare Y with the discriminant value. If the former is larger than the latter, the individual to which the bone belongs could be regarded as a male, and in the reverse case it would be a female. The probabilities of misclassification in this way were given in the last columns of tables 2 and 3.
To see whether the probabilities of misclassification given in the tables fit to the population other than our materials or not, the author has made a blind test on the skulls belonging to the Department of Anatomy, University of Hokkaido School of Medicine. These materials were 35 male skulls selected at random from 98 individuals and 4 measurements, X₁, X₂, X₃ and X₄, were made on them. Their statistical values are shown in table 4. The result was that 4 skulls out of 35 (11.4%) were misclassifid. This proportion was approximately the same as that theoretically computed. Thus it may be stated that the theoretical probabilities could be amplified on the sexing of the recent Japanese in general.
The author further discussed on the limits of this method. They are briefly summarized as follows: 1) This method of sexing is very useful for the extremity bones but its confidence coefficient is somewhat lower for the skulls, so that it would be desired to add some observations of specialist in the latter case. 2) The probabilities of misclassification was especially high in the skulls with relatively broad or low brain cases. It may, therefore, be required to observe carefully in sexing of the skulls with such shapes.
As a whole, the discriminant function, as a method of sex diagnosis, is much more useful than the other method heretofore in use in its objectivity, simplicity and accuracy.