[Reproducibility of spectrophotometric and flame photometric methods]

The reproducibility of the studied biochemical methods was determined by means of the criterion of dispersion of the results approaching the average arithmetic value--the coefficient of variance 'V'. A total of thirty analyses per series were made of one and the same sample, under one and the same conditions, by one and the same laboratory assistant. The blood plasma of cows was studied spectro-photometrically with regard to the levels of urea, total protein, sialic acids inorganic phosphorus, carotene, cholesterin, alkaline phosphatase, aldolase, GOT, GPT, and whole blood (for blood sugar). Complexonometrically, blood plasma was studied for total calcium and magnesium. Flame-photometrically, blood plasma was studied for the content of potassium and sodium; these elements were also followed up in erythrocytes and urine of cows as well as in semen of boars. Employed were methods that were routinely used within the system of the Research and Productional Veterinary Union as suggested by Tsvetkov et al. in their Manual of Methodical Guidance. The values of the coefficient of variance for the variance series of each index were compared to the admissible boundary of V (AB-V) as calculated by Tonks' method. Stated are the sources of mistakes for each of the methods tested, and possibilities are sought to optimize the values of V through eliminating accidental errors.(ABSTRACT TRUNCATED AT 250 WORDS)