The mineral assemblages from metamorphosed slightly calcic pelitic rocks of the Taconic Range in southwestern Massachusetts and adjacent areas of Connecticut and New York were studied petrographically and chemically. These rocks vary in metamorphic grade from those below the chloritoid zone through the chloritoid and garnet zones into the kyanite-staurolite zone. Microprobe data on the ferromagnesian minerals show that the sequence of increasing Fe/ (Fe+Mg) value is, from the lowest, chlorite, biotite, hornblende, chloritoid, staurolite, garnet. Hornblende, epidote, garnet, and plagioclase are the most common minerals that carry significant calcium. Biotite is persistently deficient in alkali but is abnormally rich in octahedral aluminum to such an extent that the overall charge balance can be ascribed to an AI=K+ (Fe,Mg) diadochy. Muscovite contains small though persistent amounts of iron and magnesium in octahedral positions but has a variable K/Na ratio, which is potentially useful as a geothermometer. One low-grade muscovite is highly phengitic, but the white micas in rocks from metamorphic grades higher than chloritoid zone do not contain significant phengite components. Chlorite is persistently high in aluminum and so its ratio of divalent ions to aluminum is approximately that of garnet. Many garnets show pronounced zoning in manganese and less pronounced zoning in calcium. Garnet coexisting with hornblende contains a high proportion of the grossularitic component. The calcium content is significant in all the analyzed garnets, except those from a cummingtonite-bearing sample that is free of muscovite. This suggests that in slightly calcic pelitic rocks, calcium-free garnet cannot coexist with muscovite. Most of the mineral assemblages formed in the presence of excess quartz and muscovite. The phase-petrologic analysis, made with the aid of an eight-phase multisystematic model, shows the following major points: 1. Chloritoid and staurolite coexist in a definite interval of prograde metamorphism. 2. Biotite-chloritoid does not constitute an alternative assemblage to garnet-chlorite-muscovite, because the former combination is found predominantly in the presence of the latter combination. Because the garnet contains lime, all five phases are stable together in lowlime pelitic rocks. 3. The first appearance of staurolite in the area does not correspond to the reaction leading to the first intrinsic stable existence of this phase. Inasmuch as the first appearanc,e of staurolite is always in chlorite-bearing as semblages, I suggest that the mapped staurolite zone marker corresponds to a reaction whereby staurolitechlorite becomes stable. The probable lower grade chemical equivalent, for example, chloritoid-aluminum silicate, however, has not been found in the area of study. Several staurolite-forming reactions discussed in the literature are ruled out because of the relative siderophility of the minerals. A second staurolite isograd involves the reaction, chloritoid+chlorite+muscovite= staurolite+biotite. A third isograd involves staurolite+ chlorite=biotite+kyanite; this reaction is postulated on the basis of the observed assemblage biotite-kyanitesta urolite-garnet-muscoviteplagioclase-quartzilmenite. 4. In low-grade rocks, epidote is stable considerably before the first appearance of chloritoid. The nature of the high-aluminum phase in low-grade rocks that leads to the formation of chloritoid remains obscure. The epidote is always rich in ferric iron (pistacite content of about 1/ 4 to 1/3). Garnet-bearing assemblages (with or without epidote) are formed next as metamorphic grade increases. The next more calcium-rich silicate is hornblende, and despite the meager data on assemblages that include hornblende, the first intrinsic appearance of this phase has probably been recorded. At highstaurolite grade, the most calcium-rich assemblage in pelitic rocks is hornblende-garnet-biotite-plagioclase (bytownit