Hicock, in his feasibility study on alpine drift prospecting, presents a valuable physical model for valley-controlled glaciation by describing the route travelled by the debris of the Myra Creek Glacier along the western part of the Buttle valley. His findings are well documented by clast lithology and heavymineral studies. However, when discussing the glacial dispersal of mineralized material in the till, he states "Lodgment till exposures. . .reveal a short (approximately 20 km) glacial dispersal train of Westmin massive sulphide ore in the clay fraction only. . . ." This conclusion is difficult to accept considering that no other fractions were chemically analyzed and only two other fractions (pebbles and fine-sand heavy minerals) were visually investigated. Subsequently, in his recommendations, Hicock states that "the clay (-0.002 mm) fraction is the best to analyze because it is the most representative of the preoxidation proportions of metals in the till (Shilts 1975, 1984)." He further suggests that this recommendation may apply for other alpine areas. In contrast, many geochemists tend to avoid the clay fraction altogether in sample analyses because cation exchange capacity, clay mineralogy, clay content, pH, and hydromorphic conditions all influence the absomtion of metals and affect mobile and immobile elements in different ways (C. F. Gleeson, personal communication, 1986; P. J. Rogers, personal communication. 1986). Results from a till geochemistry project in northern New Brunswick (under the Canada New Brunswick Mineral Development Agreement) suggest that the -250 mesh fraction is more suitable than the clay fraction for detecting metal anomalies. In order to study the effect of grain size, a pilot study was camed out using basal till samples taken at 2 km spacings (i.e., 1 per 4 km2). The till is generally thin over bedrock and very rarely more than 2 m thick (Pronk 1986, 1987~). The sieved -80 and -250 mesh (clay plus silt) fractions and the centrifuge-separated clay and silt fractions from approximately 300 basal till samples were analyzed for Cu, Pb, Zn, Ag, Mn, Fe, Co, Ni, As, and Sb. Although the clay fraction samples produced a higher background to anomaly contrast, the sieved fractions gave a similar number of anomalous values and outlined the same anomalous areas. Some of the clay samples produced false anomalies, caused by hydromorphic conditions at the sample site (Pronk 1987~). In addition, the low clay content of many tills in this area poses a further problem: in one study, two-thirds of the tills contained less than 1 % clay (A. A. Seaman, personal communication, 1987). In order to decide which grain size to use for geochemical analysis, a pilot study is essential. In our case, the -250 mesh