Between 1.5 and 6.5 g of dried sediment from each sample was treated with sodium pentaphosphate and H2O2 (to disperse the sediment), heated, placed in an ultrasonic bath for ~5 s, and sieved to retain only material >45 µm. Calcareous material was removed with HCl, and heating, ultrasound, and sieving were repeated twice more. Each sample was randomly distributed on gelatin-coated coverslips after the settling method of Moore (1973), using the coverslip holders of Lazarus (1994), and the coverslips were fixed to slides with Canada balsam. For each sample, two slides were prepared and at least half of one slide was examined, depending on radiolarian abundance, mostly with a Zeiss Axioskop at magnifications between 100x and 640x.
Overall abundances of radiolarians as well as those of other material, particularly pennate diatoms, centric diatoms, silicoflagellates, and lithic fragments (including ash) were noted (Table T1). The amount of dry sediment used per square millimeter of slide surface was calculated for each sample and is also listed in Table T1 as an aid to future use of the slides in quantitative analysis.
Although a broad range of species of radiolarians is present in each slide, attention was most closely paid to known stratigraphically significant species. Presence (or absence) of these species was noted in the form of estimates of their relative abundances (described as abundant, common, few, or rare; each rank being approximately four times as numerous as the next rank [Lazarus, 2001]). Most images of specimens for plates were captured digitally with a Sony black-and-white video camera with 400 lines of resolution and a Hauppage WinTV Frame Grabber Card at a 640 pixel x 480 pixel resolution. A few were captured with an Olympus DP/50 digital camera and BX51 microscope at resolutions up to 6 Mpixel. The digital pictures have been adjusted for contrast, brightness, and pixel resolution but have not been otherwise altered.