VOLCANIC PEBBLES IN THE ROCKWELL FORMATION, MARYLAND AND PENNSYLVANIA: SEARCH FOR THE LOST ARC

An unusual polymictic diamictite was deposited along a 400 km long belt from northeastern Pa. to east-central WV during the late Famennian of the Devonian (approximately 360 Ma). This deposit is in the basal Spechty Kopf Fm (NE Pa.), the basal Rockwell Fm (S Pa. and W Md.), and the top of the Hampshire Formation (WV). It contains a small percentage of exotic, igneous pebbles. In an attempt to determine their provenance, a sampling of the diamictite deposited at Crystal Spring, Pa., and Sideling Hill, Md. for volcanic and hypabyssal lithologies yielded 21 pebbles suitable for analyses. Most are distinctly porphyritic, a characteristic of convergent margin volcanics. They lack any macro-scale structural fabric. Using the alkali vs. silica diagram of LeBas, this sampling yielded 1 basalt, 2 basaltic andesites, 5 andesites, 8 dacites, and 5 rhyolites (mostly silicified). However, chloritization is pervasive suggesting that K₂O may have been lost during deuteric alteration, causing a substantial downward shift on the LeBas diagram from the original rock type. In addition, because of an intentional bias to sample only the finer-grained, volcanic-looking igneous rocks, it is likely that the more resistant rhyolites are overrepresented, and the less durable basalts, more susceptible to grinding down during transport, are underrepresented with respect to their source. Also, in the glacial environment of this deposit (Brezinski et al., 2006), andesite from the higher elevation stratovolcanoes may have been more likely to be transported than basalt from lower elevation satellite conduits. A few andesitic lithic welded tuff pebbles found at Sideling Hill provide evidence of subaerial pyroclastics. Similarity within REE_CN and other geochemical plots suggests that the source was from a few, related volcanic centers. All have Eu_CN (-) indicating loss of plagioclase and 15 have La_CN < Ce_CN. On an Hf/3-Th-Ta diagram, 20 plot in the calcalkaline portion of the destructive plate margin field. Using Ti-Zr, samples having < 62% SiO2 all plot as calc-alkaline. Cr-Y and other diagrams suggest predominantly VAB (volcanic arc basalts). Attempts to locate the source arc have met with only limited success, in part, because potential source terrains such as the James Run may be too metamorphosed at the present levels of exposure.