SAMPLE PREPARATION

To obtain clean radiolarian residues for microscopic examination, sediments must be disaggregated, sieved to remove the clay-size fraction, and acidified to eliminate the calcareous component (Sanfilippo et al., 1985). Soft sediments are commonly disaggregated after being boiled in a solution of hydrogen peroxide and sodium pyrophosphate for a few minutes, but if lumps remain after sieving through a 63-µm (or 44 µm) mesh sieve, the coarse fraction is returned to the beaker, dried, and boiled again. Successive dryings and retreatments are sufficient to clean most radiolarian samples. The cleaned residue is sieved at 150, 63, and 44 µm and pipetted evenly onto labeled glass slides. When completely dry, air is expelled from the skeletons by the addition of a few drops of xylene, Canada balsam is dropped onto them, and a 22 mm × 44 mm coverslip completes the preparation.

Normally, lithologic smear slides of sediment samples will tell the observer whether or not radiolarians are present in the sample. Identification of radiolarians in the smear slides from this leg, especially from Hole 999B, is difficult because of the great degree of lithification and the overall scarcity of siliceous microfossils. In addition, most of the samples we obtained from Leg 165 were so highly lithified that disaggregation was difficult if not impossible. The standard method of radiolarian preparation for deep-sea sediments as described above was used, but most samples required repeated treatments (five to six times) with hydrogen peroxide and sodium pyrophosphate, sometimes over a period of days. We found it necessary to make many slides (up to 10 or more) from a single sample to find a sufficient number of well-preserved forms to make a stratigraphic determination. Many of our samples were barren of radiolarians. Frequently, volcanic glass shards and their alteration products, such as smectite, remain in the siliceous residues as minor constituents along with feldspar, opaque minerals, and traces of pyrite (P. Worstell, pers. comm., 1998).

NEXT