I looked around a little bit. Found this
http://www.springerlink.com/content/q635t364q522q64u/. It may be of interest. Most of it is over my head, but from what I understand, formation processes differ from region to region - and even within a region, hence the relevance of "different veins" we often see discussed here on the site. Moreover, it would appear that the N. American coticule's crystal is bell shaped and not a rhomboid like it's Belgian cousin - and this may very well affect the way it would function as a whetstone or hone. Here is the abstract:
Crystal size distribution (CSD) in rocks and the kinetics and dynamics of crystallization
Katharine V. Cashman and John M. Ferry
Crystal size distributions (CSDs) measured in metamorphic rocks yield quantitative information about crystal nucleation and growth rates, growth times, and the degree of overstepping (DeltaT) of reactions during metamorphism. CSDs are described through use of a population density function n=dN/dL, where N is the cumulative number of crystals per unit volume and L is a linear crystal size. Plots of ln (n) vs. L for olivine+pyroxene and magnetite in high-temperature (1000° C) basalt hornfelses from the Isle of Skye define linear arrays, indicating continuous nucleation and growth of crystals during metamorphism. Using the slope and intercept of these linear plots in conjunction with growth rate estimates we infer minimum mineral growth times of less than 100 years at DeltaT<10° c,="" and="" nucleation="" rates="" between="">–4 and 10–1/cm3/s. Garnet and magnetite in regionally metamorphosed pelitic schists from south-central Maine have CSDs which are
bell-shaped. We interpret this form to be the result of two processes: 1)
initial continuous nucleation and growth of crystals, and 2)
later loss of small crystals due to annealing. The large crystals in regional metamorphic rocks retain the original size frequency distribution and may be used to obtain quantitative information on the original conditions of crystal nucleation and growth. The extent of annealing increases with increasing metamorphic grade and could be used to estimate the duration of annealing conditions if the value of a rate constant were known. Finally, the different forms of crystal size distributions directly reflect differences in the thermal histories of regional vs. contact metamorphosed rocks: because contact metamorphism involves high temperatures for short durations, resulting CSDs are linear and unaffected by annealing, similar to those produced by crystallization from a melt; because regional metamorphism involves prolonged cooling from high temperatures, primary linear CSDs are later modified by annealing to bell shapes.
but it might shed a glimmer of light on the general nature of coti's, especially for those of you more learned than I.