Clinical chemical reagent tablet weight distribution.

T1HE1tE APPEAR TO BE virtually no published studies OH the weight (list ributiOllof tablets intended for clinical chemical purposes. Studies 011 the output of a machine, considered as a frequency distribution, are commoii today in industry. Tablets, made with machines on which the tablet weight may be adjusted, are cOmmOllly used in manufacturing pharmacy (1). Hand-operated tablet presses are available, but are not yet used widely in chemical laboratories. It was thought that an investigation of weight distribution of reagent tablets supplied by an experienced manufacturer would l)e of value in showing tile derree of reproducibility among individual tablets. This reproducibility then would give a measure of tile disjJellsilig accuracy to i)e expected when using reagents in tablet torm. The tablets selected for this study were Warner-Chilcott PhosphatabsAlkaline Quantitative, a mixture of a sodium salt of pllellolphthalein hphosphate with tris(hydroxymethyl)aminomethane buffer and, pos. sibly, a binding agent (2). The tablets taken from 2 bottles (Lots 3268106 and 2980096) were individually weighed to an accuracy of ±0.1 mg. A total of 121 tablets taken at random from the bottles were weighed. A graph was made on arithmetic probability paper (3) and is shown in Fig. 1. The straight line indicates that the distribution is gaussian or normal. inspection of Fig. 1 shows that tile mean weight was 21.8 mg. Tile