It is believed that the Chesapeake structure, with its age of 35.3 million years, is one of the most recent large impact crater on earth. However, seismic investigations and some "flat" boreholes in which breccias were drilled are not sufficient to confirm initial assumptions.
In 2005, the final proof was to be provided by a deep borehole in the center of the structure. As part of the International Continental Scientific Drilling Program (ICDP), the US Geological Survey induced the core drilling Eyreville, with a depth of 1766 m.

According to the scientific evaluation of the drill cores, it is believed to now have the confirmation that the Chesapeake Bay structure actually resulted from an impact event. This can be doubted, however, because once again was ignored a clearly existing magmatic melt. It was not recognized that the "melt-rich suevites" found in a depth range of 1397-1551 m are, in fact, breccias infiltrated with a dacitic-rhyolitic magma.

This is the result by my investigation of the many analyzes (BARTOSOVA et al.) of melts in this interval. They confirm a dacitic-rhyolitic composition of the pure melt. A wide range of deviating compositions of melts are the result of a mix of the dacitic-rhyolitic melt with molten breccias. A special feature are pure silica fractions (with very high SiO2 contents), as well as the cristobalite and tridymite crystals grown during the cooling.

          
Pure dacitic-rhyolitic melt with pyroxene (+plagioclase phenocrysts)

Pure melt with Silica bodies (Ballen quartz)

The center of the Chesapeake Bay impact structure was drilled during 2005⁄2006 in an ICDP-USGS drilling project. The Eyreville drill cores include polymict impact breccias and associated rocks (1397–1551 m depth). Tens of melt particles from these impactites were studied by optical and electron microscopy, electron microprobe, and microRaman spectroscopy, and classified into six groups:

• m1— clear or brownish melt,
• m2— brownish melt altered to phyllosilicates,
• m3— colorless pure silica melt,
• m4— melt with pyroxene and plagioclase crystallites,
• m5— dark brown melt,
• m6— melt with globular texture.

Type m1 is a clear or brownish melt, relatively homogeneous, slightly altered, with totals of ~87 wt%, and ~77 wt% SiO2.
Type m2 is a brownish melt, totally altered to phyllosilicate minerals, inhomogeneous, with abundant undigested clasts, and with low totals of ~80 wt%, and ~56 wt% SiO2.
Type m3 is a colorless melt with some brownish stains, with >95 wt% SiO2 and totals close to 100 wt%.
Melt type m4, with pyroxene and plagioclase crystallites, forms matrix in the impact melt rocks and has approximately 75 wt% SiO2 and totals close to 100 wt%.
Dark brown melt particles of type m5 have commonly undigested clasts, only 54 wt% SiO2, and the highest contents of Al2O3 and FeO; totals are approximately 91 wt%.
The last melt type m6, brownish melt with typical globular texture, occurs exclusively in the upper impact melt rock interval; it has 75 wt% SiO2 and the totals are approximately 96 wt%.

The melt particle types exhibit rather distinct, yet partly overlapping chemical compositions. However, compositions are quite variable among melt particle types and individual melt particles of one type. Particles display primary compositional differences (e.g., schlieren) due to mixing of melt phases with different composition, and irregular distribution of crystallites and undigested clasts. There are also secondary changes due to hydrothermal alteration—the melt particles were altered to phyllosilicate minerals, and secondary zeolites and anatase formed. Parts of the silica melt recrystallized to ballen quartz, and rare ballen cristobalite was also noted. Except for the nearly pure silica melt (melt type m3), no mono-mineralic melts were found. The chemical analyses and mixing calculations suggest that many of the melt types are relatively silica-rich. The pre-event impact sedimentary formations (e.g., those similar to rocks from the Cretaceous age) seem to have been main components for the melt types m1, m4 and m6.
The basement-derived schist or gneiss could also be a precursor for some melt types (namely the most abundant melt types m2 and m5) according to the calculations. However, microscopic observations of some partly melted clasts and comparison of chemical composition of the melt particles and sedimentary clasts from the Eyreville drill cores suggest that melt types m2 and m5 possibly formed from a fine-grained, clay-rich sediment. Melt type m3 is melted quartz, quartzite, or quartz arenite.
We suggest that the pre-event impact sediments are the most important precursors for the identified melt types. The sediments constituted a large part of the target (1 km thick) and were probably largely melted or vaporized. However, according to the diagrams and mixing calculations and given the dimensions of the crater, a crystalline basement (probably schist ⁄ gneiss) precursor could have been involved in formation of some melt particles. The composition of the melt particles is highly variable and does not indicate a uniform homogenized melt source.
Alteration could have substantially changed the composition (by, e.g., leaching of cations) of most melt types and makes the estimation of the melt precursors difficult. The Eyreville drill cores provide only a limited suite of samples for such a large impact structure and it is possible that other types of melt particles or even larger melt bodies are present in other parts of the Chesapeake Bay impact structure.

Monoclinic tridymite in clast-rich impact melt rock from the Chesapeake Bay impact structure
J.C. Jackson, J. Wright Horton, I-M. Chou, H. E. Belkin; American Mineralogist (2011)

X-ray diffraction and Raman spectroscopy confirm a rare terrestrial occurrence of monoclinic tridymite in clast-rich impact melt rock from the Eyreville B drill core in the Chesapeake Bay impact structure. The monoclinic tridymite occurs with quartz paramorphs after tridymite and K-feldspar in a microcrystalline groundmass of devitrified glass and Fe-rich smectite. Electron-microprobe analyses revealed that the tridymite and quartz paramorphs after tridymite contain different amounts of chemical impurities. Inspection by SEM showed that the tridymite crystal surfaces are smooth, whereas the quartz paramorphs contain irregular tabular voids. These voids may represent microporosity formed by volume decrease in the presence of fluid during transformation from tridymite to quartz, or skeletal growth in the original tridymite. Cristobalite locally rims spherulites within the same drill core interval. The occurrences of tridymite and cristobalite appear to be restricted to the thickest clast-rich impact melt body in the core at 1402–1407.5 m depth. Their formation and preservation in an alkali-rich, high-silica melt rock suggest initially high temperatures followed by rapid cooling.

Petrology of impact melt rocks from the Chesapeake Bay crater
A.Wittmann, R. Schmitt, L. Hecht, D. Kring, U. Reimold, H. Povenmire; Geological Society of America, Special Paper 458 (2009)

In drill core Eyreville B, the section of pre-resurge impactites (depth of 1397-1551m) includes bodies of clast-rich, unbreccieted impact melt rock of depths of 1402-1409.5 m as holocrystalline and at 1450-1451.5 m as hypocrystalline varieties. These bodies contain more massive impact melt rock in their central parts, from 1450 to 1451.5 m (M1), and between 1402 and 1407.5 m (M2).
Hypocrystalline clast-rich impact melt rock (M1):
Macroscopically, M1 is composed of fluidal-textured, banded, clear to brown melt (30-70vol%) that entrained lithic clasts. Dark, fluidal streaks in the rocks are partly fused remains of the fine clastic debris or mafic components that were mixed into the melt. Microscopically, the melt exhibits some glassy, clear domains that are isotropic and intergrown by zoned orthopyroxene crystals that exhibit single-lath and dendritic and euhedral, spinel crystals. In clear melt streaks, brown, spherolitic aggregates of radially intergrown aluminosilicate microphenocrysts are present. Dark, aphanitic streaks of melt contain euhedral Fe-Ti-O crystals, few fan-like plagioclase phenocrysts, and small, equant quartz clasts in a scmectitic groundmass.
Most quartz clasts exhibit a dark, bubbly contact aureole against the melt but no halo of pyroxene crystals. Deplectic quartz glass was recrystallized to polycrystalline  ballen-textured quartz. The Raman spectroscopy indicated the presence of alpha-quartz.
Holocrystalline clast-rich impact melt rock (M2):
This mostly holocrystallin, clast-rich, banded, massive gray impact melt rock is 5.5 m thick. Dark streaks are drawn out into discrete strings and indicate little mixing with the typical gray melt matrix. They are composed of similar components as the dark streaks in M1, except for the smectite  matrix, which is mostly overgrown with feldspar crystals. Quartz clasts are generally subrounded, have a brownish tint, and frequently display decorated PDFs. The M2 samples exhibit similar phenocryst phases in the melt as M1. The melt rock matrix is fully crystallized.
Electron microprobe data of the melts:
The composition of clear, glassy melt was analyzed in pristine domains of sample W84 from M1. The  results indicate a peraluminous, rhyolitic composition with a water content of ~5%. The  crystallized melt matrix in sample W61 from M2 indicate a similar geochemical affinity.
Dark streaks in the melt rocks of M1 and M2 were analyzed. Analyses indicate a reservoir for MgO, Al2O3, Fe and TiO2, whereas SiO2, Na2O3 and K2O are dramatically depleted in the dark streaks compared to the composition of the glassy, clear melt.

The ICDP-USGS Deep Drilling Project in the Chesapeake Bay Impact Structure: Results from the Eyreville Core Holes
Ch. Koeberl, K. Miller, U. Reimold; Geological Society of America, Special Paper 458, 2009