The preference of the authors of the quantitative igneous rock classification for an artificial rather than a natural system, coupled with their invention of a new nomenclature to accompany the classification, indicates that some essential elements of scientific work are not empirically ascertained but are proposed and accepted (or rejected) by the relevant scientific community as a matter of free choice. The use of igneous rocks as exemplars in the education of novice geology students is discussed. It is claimed that the CIPW classification could not have been produced by a single individual geologist. The factors that allowed for the collective success in the creation of the quantitative classification are examined.Upon publication of their monumental quantitative chemico-mineralogical classification (CIPW 1902, 1903), C. W. Cross, J. P. Iddings, L. V. Pirsson, and H. S. Washington immediately received numerous letters of congratulation. Initial published reviews ranged from highly supportive to suspicious. To help buttress their classification, Washington (1903) published a compilation of igneous rock chemical analyses and Iddings (1903) published several diagrams to drive home the point that a natural classification of igneous rocks was not feasible. Led by Washington, Pirsson, and Cross, several geologists began using the CIPW classification in their petrological studies and some contributed new sub-rang names. In the meantime, Iddings worked on the first volume of a projected two-volume work on igneous rocks based on the quantitative CIPW scheme. Unsympathetic to artificial, overly precise classifications, Harker in particular rejected the CIPW system and its norm calculations and European geologists generally were unenthusiastic. Cross (1910b) offered a major rebuttal to the criticisms, particularly those of Harker, in which he challenged the likelihood of producing a valid natural classification of igneous rocks. Iddings (1913) published the second volume on igneous rocks in which he developed an elaborate correlation between the old qualitative system and the new quantitative CIPW scheme. Washington and Pirsson produced many more petrological studies of Mediterranean volcanic rocks, New Hampshire, and Hawaii that incorporated the quantitative system. Washington (1917) produced a vastly expanded compilation of chemical analyses arranged in accord with the CIPW system. Criticisms, however, continued to mount from Fermor, Daly, Shand, and others, while Tyrrell and Johannsen were lukewarm toward the new classification. The criticism that the CIPW system was of little value in fieldwork repeatedly surfaced. Dissatisfaction with the quantitative scheme led to the publication of many new classifications by geologists, such as Hatch, Winchell, Lincoln, Shand, Holmes, Johannsen, and Niggli. With the creation of satisfactory quantitative mineralogical classifications, the increasing ability to determine the proportions of minerals quantitatively, and the death of Iddings and Pirsson, enthusiasm for the CIPW system gradually began to wane. By the 1960s the classification had become a thing of the past. The value of the norm calculation, however, gained recognition and has survived to the present, assisted no doubt by the capability for doing the necessary calculations by computer.
