The Chelmsford granite is quarried in and around Oak Hill, about six miles west of Lowell, Massachusetts. The granite‐area is about eight miles long and one to three miles wide, and its longer dimension has a northeast bearing which is parallel with the regional axis of foliation in the country rock. The writer favors a hypothesis that much of the granite represents granitization of biotite schist under hydro‐magmatic invasion, and under stress that was insufficient to deform or distort the preexisting country rock‐structures to any marked degree. Banding of the granite is due chiefly, to variable amounts of biotite. The bands are thin and remarkably uniform. The banding, attitudes of schist inclusions, and distribution of the entire mass con form to the regional structural elements. Lit‐par‐lit injection‐bands are not clearly discernible although many of the light‐colored bands are doubtless of such origin. The absence of flow‐schlieren is noted. Petrographic evidence shows a remarkable amount of metasomatic fabric. Undoubted primary granitic texture is practically wanting. There is a high degree of orientation of biotite, muscovite, and groups of small mosaic quartz‐grains. The contacts of inclusions with adjacent granite‐bands are very irregular and interpenetrative, and granitic material also penetrates intricately along schistosity planes within the inclusions. Large irregular feldspar porphyroblasts and strings of such grains are distributed through the inclusions, generally with decreasing concentration toward the centers. Pegmatite and quartz veins, some of which carry much tourmaline, are abundant in some of the quarries. Crystallization zoning of feldspars is very rare. Feldspars are of three generations: (1) Early polysynthetically twinned porphyroblasts and irregular groundmass grains of oligoclase or albite‐oligoclase; (2) microcline and orthoclase, chiefly as porphyroblasts, and in part replacing the earlier plagioclase; and (3) albite and albite‐oligoclase, untwinned and mottled, as individual crystals, streaks, patches, and porphyroblasts, replacing microcline, quartz, and mica. Incomplete twinning, corrosion‐contacts, optical anomalies, inclusions of minerals, pseudo‐perthitic and myrmekitlc grains are among the diagnostic petrographic features that contribute to the theory of replacement.