Alba Regia. Annales Musei Stephani Regis. – Alba Regia. A Szent István Király Múzeum Évkönyve. 26. 1989-1992 – Szent István Király Múzeum közleményei: C sorozat (1997)

Tanulmányok – Abhandlungen - Bartosiewicz László: A Székesfehérvár Bestriary: Animal Bones from the Excavations of the medieval City Wall. p. 133–167.

using body measurements taken on 175 living animals. The comparative sample included Hungarian Fleckvieh and dairy Holstein Friesian cows as well as the second generation of their crossing which has been upgraded by the dairy breed (75 % Holstein Friesian). During previous zoological research the slenderness of the front leg was found to decline along with an increase in live weight. In order to make this tendency measurable, an index was constructed using the shank circumference as numerator and the difference between withers height and chest depth as denominator It was assumed that the aforementioned proportion is analogous to an index using four skeletal measurements. These include the smallest breadth and smallest depth of the metacarpus and the sum of the metacarpus and radius greatest lengths. The first two measurements were used in calculating the circumference of an ellypse which supposedly corresponds to shank circumference in the numerator of the previous index (times the sum of smallest breadth and smalles depth). The summarized length of the metacarpus and radius were used instead of the extremity length defined by the difference of the withers height and chest depth in the denominator of the prevoius formula. Obviously, the bone measurements yielded smaller values than the body dimensions included in the first index. It was assumed, however, that the actual values of the two fractions differ to a much smaller degree from each other. Finally, as a research hypothesis, a similar growth rate for the two indices was presumed as the function of live weight. That is, a more intensive relative increase in the index value itself, expressing a decline in the slenderness of the extremity that parallelled increasing live weight. It is well known that most of the animal growth tendencies (especially those in mammals) may be modelled by a cumulative curve which can be transformed into a linear form using logarithmic scales (Brody 1927; Fábián 1969). Thus, the freshly estimated greatest length of radius was entered into a regression analysis using the logarithm of the index as a dependent variable and the logarithm of the cubic root of live weight as an independent variable. This last transformation was introduced in order to avoid bias resulting from dimensional distortions: The indices were constructed from linear measurements while live weight displays a change which is three dimensional in character (Halstead-Middleton 1972, 21). Results of these calculations are summarized in Table XXIV. The live weight obtained this way corresponds to that of a Lebedin Brown cow from the Ukraine or to the less common Aubrac breed from Southern France (HornDohy 1970, 33-103). Due to the low but highly significant correlation between the live weight and the calculated index in the comparative sample containing three genotypes, the standard error of the estimate is impressively small. In fact, natural fluctuations of live weight are much larger due to seasonal nutritional differences, occasional disease, stage of gestation etc. In order to complement information on withers height and live weight a third trait, fat free carcass weight, was estimated-from the lateral and medial lengths of the astragalus and the distal breadth of the same bone. Comparative data used in this part of the calculations come from an elaborate study by Noddle (1973). One of he bivariate scattergrams suggested a linear relationship between the multiple of the three astragalus measurements and the proportion of this composite value to fat free carcass weight. A regression analysis carried out along the lines of the previous two estimations in this study showed a highly significant correlation between these two variables based on data from thirteen cows. Four of these belonged to the South Devon breed while three represented the first generation of a Hereford X Aberdeen Angus crossing. The comparative sample also included two Charolais X Jersey crosses, while the rest of the cows were Friesians. In spite of the relatively great standard error of the estimate obtained for the JM/1 cow, fat free carcass weight is much less influenced by environmental effects than live weight. The apparent inaccuracy is rather due to the small size of the comparative sample than to anatomical variability. According to the mean estimate the Turkish period cow reconstructed in this study seems to be most similar to a Hereford X Aberdeen Angus cow in Noddle's list in terms of this characterictis (Table XXV). When estimates for the three traits are compared from a quantitative point of view, they may be summarized as being similar to the type that is known as brown cattle. Although astragalus measurements suggest a relatively low fat free carcass weight (comparable to that of small stature beef breeds) this apparent underestimation may reflect the possibly smaller proportion of meat output to body size in the Turkish Period specimen. Caprines The small Ruminant bone sample is usually difficult to sub-divide into remains of sheep and goat due to the great osteomorphological similarity between the remains of these two species. Although the morphological distinction between a relatively few skeletal elements is possible (Boessneck et al 1964), the results in many cases must be complemented by additional informatioin on differences between habitat preferences (Bökönyi 1974, Uerpmann 1982), behavioral characteristics (Shackleton-Shank 1984) and the ways these two species have been exploited. Non-critical use of such diachronic analogies for comparison, of course, may lead to the overestimation of meat yields in medieval cattle which, as a less improved form, probably had a smaller dressing percentage than the average cattle contemporary to our times. 137