In spite of the amount which recent publications have added to our knowledge of Roman building methods and materials, there is still some inaccuracy to be noted in the use of terms due very often to a misconception of the chemical constituents of the materials in question. In the present article will be discussed the confusion which still exists between the words sand and pozzolana. In Daremberg and Saglio, s.v. arena, Guillaume states that in the composition of hydraulic cements, “le sable fossile, noir, blanc ou rouge preféré par Vitruve, est probablement la pouzzolane.” In the recent articles by Miss Van Deman on Roman construction, one finds frequently the term pozzolana-arena, and the statement is made that “mortar is composed of pozzolana, which is called by Vitruvius arena fossicia, and lime.” In the new translation of Vitruvius, by J. Prestel, while the word pulvis is correctly translated in the text as “eine staubartige Erdmasse,” in the notes pulvis is translated as “sandreiche Erde.” Moreover, further on in the text, this same pulvis is described as a kind of sand, “Sandart.” Finally may be mentioned the reference in Delbrück's Hellenistische Bauten in Latium where it is stated that in the construction of opus caementicium, the Romans employed a “sharp volcanic sand.” This I shall hope to show to be a contradiction of terms, as “sharp” sand has nothing to do with volcanic activity.