The lacustrine carbonates of the Bonney Drift, Taylor Valley, Antarctica were deposited in lakes proglacial to glaciers advancing through the valley, and now occur as lag and clasts in the glacial drift formed by the glacier. They commonly contain 8 to 25% detritus, including high levels of silica, titanium, iron, manganese, and other elements; making the extraction and purification of uranium and thorium problematic. A standard U-Th dating process was adapted to deal with the technical problems peculiar to the samples dated. Silica often formed a gel at the head of the U-Th separation column, and was removed using HF. Titanium frequently precipitated out during the conversion of the thorium fraction from a HCl matrix to HNO₃ form (used for further ion exchange purification) causing very low thorium recovery. This was overcome by using much higher volumes during conversion. The uranium eluate from the separation column often contained high levels of iron and manganese. These were removed on a small ion exchange column in HNO₃. The resin and wash volumes were adjusted to allow iron and manganese to wash off, but allowing the uranium to be held and eluted subsequently. An electrodeposition method was developed to achieve the thin plates necessary for alpha spectroscopy. Silver cathode discs were used in a teflon plating cell, with a stirrer rotating on the spiral planar platinum anode. An NH₄Cl/HCl electrolyte was used, adjusted in pH after sample uptake. Of the 62 samples successfully dated, 13 were eliminated for such reasons as low resolution and spectral shift; leaving 17 duplicates, one triplicate and 12 single analyses for further consideration. Samples with low ²³²Th/TIFFS Th levels were corrected for detrital ²³²Th addition. The dates obtained were considered in terms of age, geographic distribution, and uranium isotope ratios and concentrations. δ180 suggested that samples were derived from either Taylor Glacier meltwaters, or possibly Ross Ice Sheet Stage 6 meltwaters, and showed that δ180 values in the meltwaters increased towards the middle of each event. Three major events were identified. (i) an extensive period of deposition around 240 to 300kA (Stage 9?), This may possibly represent two events, in a lake(s) most likely formed proglacial to an expanded Taylor Glacier. (ii) a short period of lacustrine conditions around 175 to 190 kA ago, correlated with either Taylor Glacier expansion at the end of Stage 7, or Ross Ice Sheet thickening at the start of Stage 6. (iii) a period of carbonate deposition from 130 to 80 kA( Stage 5), more extensive than that at 175-190 kA. The expanding glacier (Taylor II) then advanced through the carbonates, and those of the earlier events, and redeposited. The events interpreted appear to correlate with global interglacials, as suggested by Hendy et al (1979). It seems that during interglacials, the East Antarctic Ice Sheet expands (the absence of a sea ice apron allows increased precipitation). As a result, expansion of the East Antarctic Ice Sheet would be expected as a consequence of greenhouse gas warming, offsetting sea level rises caused by melting of temperate glaciers and ice sheets.