Previous studies have shown clear differences between the anisotropies of low‐field magnetic susceptibility (LFS) and of anhysteretic remanent magnetization (ARM) in samples of the Ordovician Trenton limestone from Trenton Falls, NY. ARM was found to be significantly more anisotropic than LFS, and to exhibit a well‐defined east‐northeasterly lineation which was not apparent in anisotropy of magnetic susceptibility (AMS). Because of significant local variation in mean LFS (by approximately a factor of two) at this locality, it has been possible to reanalyze the AMS of these samples by the method of Henry , which in some cases allows mathematical differentiation of the ferromagnetic and paramagnetic susceptibility tensors. The ferromagnetic susceptibility tensor obtained in this way shows the same tectonic shortening apparent in the ARM anisotropy, but surprisingly, it shows no evidence of horizontal foliation. The paramagnetic tensor is horizontally foliated, with no lineation. The difference between the ferromagnetic susceptibility and ARM tensors suggests different ferromagnetic sources of ARM and LFS.