Matskási István (szerk.): A Magyar Természettudományi Múzeum évkönyve 100. (Budapest 2008)
Bernert, Zs.: Data for the calculation of body height on the basis of extremities of individuals living in different historical periods in the Carpathian Basin
It can be observed that practically the same tibia length belongs to the same male and female femur length (Table 3). Hence, the length of the lower extremities of males and females with the same femur length is practically the same. Based on recent Hungarian body heights and averages of lower extremities (Table 4, EIBEN et al. 1991), it can be calculated that females are 1.016 times taller then males if the length of the lower limb is identical. When comparing sitting heights, it can be observed that the upper body of females is 1.020 times longer than that of males. When calculating correlations below, I used this ratio. Finally, I prepared the linear regression equations resulting in body height data as follows (Table 6): Table 6. Formulas for estimating the body height of males Females 0.323 x Ml humerus + 68.771 0.409 x Ml radius + 73.758 0.192 xMl femur + 85.570 0.248 x Ml tibia + 81.101 0.108 x(Ml femur +Ml tibia ) + 83.621 One method for evaluating individual height data is to classify them. This is the most transparent way of data processing. We can notice in several works involving stature estimates following MARTIN'S classification that the proportion of tall individuals (tall-medium or very tall, depending on population) is extremely high. This can be observed both in case of males and females of large series. The deviation from the theoretical Gauss curve may refer to the emergence of taller populations or the selective effect of the environment. However, the cause of this phenomenon is methodological and not biological. When choosing even stature categories, we can expect the theoretical distribution even historical populations if the number of individuals in the sample is appropriately high. I took the example of the distribution of more than three thousand male femurs according to length (of people living in the Carpathian Basin in different historical periods). I chose categories of 5 millimetres from 370 mm to 539 mm (34 categories) (Fig. 4). Fig. 4 clearly shows that the distribution of length (and the stature calculated from that) follows the Gauss curve if the sample is large enough. On the basis of femurs in this figure, by using the height estimation method of SjOVOLD worked out for both sexes and different populations and assigning the calculated heights to the categories according to MARTIN'S classification, we receive the following distribution (Fig. 5). Again, the figure clearly shows the phenomenon already mentioned, namely that the distribution of stature categories does not follow the distribution of extremities, i.e. the proportion of tall individuals is too high. The explicit disharmony of distribution according to stature categories and the distribution of limb bones according to length is quite problematic because historical populations examined cannot be evaluated and compared on the basis of the distribution of stature. At this point, I would like to refer to the fact that the estimation of stature is important among others because of the transparent comparison of historical populations.
