The statement that nodule formation, especially morphology, can be predicted by the Boudreau diagenetic model led us to state that ''diagenetic models that claim to predict nodule growth rate and morphologies from elemental concentrations (1) are not verified by our data'' for Second Connecticut Lake (SCL), New Hampshire (2). Comments on metal gradients and elemental concentration of manganese (Mn) and iron (Fe) required for accretion and morphologic expression of ferromanganese nodules are in Boudreau's original paper (3). Yet, in his letter (4), Boudreau writes that elemental (metal) concentration is not discussed (1, 3).Nodule morphology and growth rates depend on water velocity, concentration of Mn 2ϩ (or Fe 2ϩ ) in the flow over the nodule, and boundary layer thickness, according to Boudreau (1, 3). If true, and we do not dispute his model for localities he studied, nodules under the same temporal and limnological conditions should develop the same morphologies. Nodules in SCL are intermixed: All morphotypes we described are closely spaced (0-10 cm apart). Pavement nodules (figure 3b in ref. 2) are intermixed with both nucleated domed-plate concentric ring (figure 3 a and d in ref. 2) and massive lattice nodules (figure 3c in ref. 2); no variations in water flow could occur within this limited spatial distribution. No underflows into the lake were detected that could generate variation in the parameters of Boudreau's model although they were sought.Boudreau asserts that nodules develop a discoidal shape that results from ''vertical diffusion'' of metal ions in solution that accrete on the leading edge of the nodule, which develop faster than their downstream portion (1, 3). Although this might explain a field of asymmetrical discoidal shapes, the literature is replete with those that occur as concentric symmetrical spheres, or massive nodules that lack concentric ring patterns (3). In SCL, some nodules grow in a concave down position in situ (50 mm deep) with an average nodule thickness of 4 mm and no competent ''front'' to which the metal ions might collect as shown in the Boudreau model (figure 20 in ref. 1; figure 3 in ref. 3). Nor does it account for our discovery of nodules that accrete concentric ring cylindrical chimney structures that range from 50 to 170 mm in height.We examined the nepheloid layer that engulfs the SCL nodules by phase contrast and Nomarski differential interference light microscopy. This mucoid layer is teeming with live motile rods, short filamentous and coccoid bacteria, and, in some cases, their sheaths both with or without trichomes. Some, inhabited or not inhabited, appeared yellowish-brown in color and are most likely associated with metal oxides. Because our purpose was to describe a new site of large, young lacustrine nodules, these observations were beyond the scope of the announcement paper (2).Models that do not consider the direct precipitation of oxides by organisms to explain proximal complex nodule morphologies ignore a large literature on microbialites (5). Biologic...