Before using the TAS or any other classification, some particular guidance by Johannsen (1937) should be kept in mind. :
Many and peculiar are the classifications that have been proposed for igneous rocks. Their variability depends in part upon the purpose for which each was intended, and in part upon the difficulties arising from the characters of the rocks themselves. The trouble is not with the classifications but with nature which did not make things right. … Rocks must be classified in order to compare them with others, previously described, of similar composition and appearance. If this cannot be done on a genetic basis, then an artificial system must answer in order to serve as a card index to rock descriptions. Although this may be an evil thing, it is, at least, the least of several evils. The TAS classification cannot be applied to all volcanic rocksas discussed in detail by Le Maitre et al (2002). Certain rocks cannot be classified/named using the diagram. For others, additional chemical, mineralogic, or
textural criteria must be used, as e.g.
lamprophyres. The TAS classification should be applied only to rocks for which the mineral mode analysis cannot be determined. Otherwise, a scheme based on mineralogy, such as the
QAPF diagram, or one of the other diagrams available for
igneous rocks may be suitable. Before classifying rocks using the TAS diagram, the chemical analyses must be recalculated to 100% excluding water and carbon dioxide. ==The TAS diagram==